Search

Breast cancer is the second leading cause of death in females. As such, women have high incidence and mortality rates of breast cancer. The incidence rate has been on the rise over time. The earlier breast cancer is caught, the better it shows prognosis and the lower the mortality rate is. For this reason, many researchers and medical doctors have heeded a lot of attention to the CAD systems to detect and classify breast cancer. They have proposed a myriad of methods and techniques. Among them, the CAD system based on artificial intelligence (AI) can process plenty of information fast, and its performance is evaluated to be high. As an AI algorithm, YOLO has excellent detection performance and can detect objects effectively in real time. In this paper, we proposed an anomaly detection model of mammography using a YOLOv4-based histogram. In terms of breast cancer diagnosis, mammography features a fast diagnosis time and an inexpensive cost. For this reason, it is often applied to breast cancer diagnosis. Mammography, however, generates an image only with brightness values, so that a mammogram image has a lot of noise and image edges are dim. To enhance these image edges, we create a difference through histogram and brightness range control and threshold-based region removal methods and expand the single channel of mammogram images using the generated images. Through the expansion, the image edges are enhanced and converted into a single channel again and are learned through YOLO. For performance evaluation, the method proposed in this study is compared with ResNet18, ResNet50, GoogleNet, and VGG16. According to an experiment, the proposed method had the highest accuracy, or 95.74%, followed by GoogleNet (89.9%), VGG16 (88.93%), ResNet50 (87.77%), and ResNet18 (87.67%) in order.

Multimedia provides a lot of information for users through various forms of information contents and information processing. These days, it is utilized and converged in diverse fields. Particularly, in the convergence of movie or TV drama media and information technologies, how to visualize or predict emotional changes on the basis of a variety of multimedia information has steadily been researched. Based on emotional changes, it is possible to analyze genres of movies or TV dramas. Viewers try to select a movie on the basis of their preferred storylines. In other words, users select and personalize the genres fitting their sentiment in the emotional flow of videos so that they want the customized media reflecting their emotional flow to be recommended. A typical method of predicting emotions shows the high accuracy of emotion prediction when text-based lines are utilized. Nevertheless, when subtle emotions in text lines are presented with voice and video information, it is possible to increase the accuracy of emotion prediction. Therefore, this study proposes a system that is capable of predicting emotional context primarily with the use of a text, and then with the analysis of emotions in the dialogs of characters or their voices in scenes, or images of characters at the point of time. In order to predict emotions efficiently, the proposed system analyzes the time information on a character’s particular emotion words from text data, extracts voice signals of the time section and converts them into a spectrogram, and saves the face image at the point of time. This imaged spectrogram and the face image for facial expression analysis are used as input data of CNN (convolutional neural network) and are trained. An emotion found in a particular paragraph is predicted, and therefore an emotional flow and an emotion in a particular scene are discerned. The developing multi-modal emotion prediction system collects the estimated emotions from convergence media and active contents (text, voice, and image) and active contents such as text, voice, and image. Finally, it predicts emotions.

Anytime, anywhere, anyone can collect not only personal information but also surrounding situation information from home or a vehicle on the move without being restricted by time and space. To this end, the need for various devices such as biosensors, wearable devices, and temperature sensors is increasing. In this circumstance, there are many problems with the interaction of multiple devices and data processes, and there is a lack of structured or unstructured emerging contextual data. Accordingly, it is necessary to come up with a data collection method to obtain a variety of information efficiently. Such a method can obtain a diversity of information consistently with the use of data convergence. Since the method makes possible decision making, entity identification, and context prediction, it is applicable to diverse areas, including object detection and context awareness. The data generated through convergence makes it possible to the improve performance of data analysis through machine learning and deep learning. For this reason, a variety of modal types are applied to data analysis. Modal convergence for analysis can improve analysis performance more than unimodal. In this study, propose the accident risk prediction model based on attention-mechanism LSTM using modality convergence in multi-modal. In order to predict the risk of an accident, the proposed method makes the convergence of structured and unstructured modality data on the basis of judgmental fusion and statistical fusion and generates data sets. It is capable of solving the problem of data shortage and obtaining a variety of information consistently. In the proposed method, firstly, preprocessing is performed, and data sets of accident risk information are generated with the uses of the road risk equation and accident risk equation. Next, the correlation coefficient and regression coefficient for each variable is calculated through regression analysis modal and correlation coefficient analysis modal. A correlation coefficient is used to judge which variable needs to be learned intensively in order for the attention-mechanism of the accident risk prediction model. A regression coefficient is used as attention-weight to adjust a data weighting factor in attention-mechanism LSTM. Accordingly, whether there is the risk of an accident is judged. Performance evaluation is conducted in two ways. Firstly, epoch-based loss is evaluated in the comparison between the attention-mechanism LSTM-based accident risk prediction model with convergence modal and the model without it. In short, their RMSE is evaluated and compared. Secondly, the proposed model is compared with conventional models in terms of RMSE.

Big data has a variety of data types, including image and text. In particular, image data-based research on face recognition and objection detection has been conducted in diverse areas. Deep learning needs a massive amount of data for learning a model accurately. The amount of data collected is different in each area, and thus it is likely to lack data for analysis through deep learning. Accordingly, it is necessary a method of learning a model effectively and predicting a result accurately with the use of a small amount of data. Also, captions and tags are generated to obtain image information. In the case of tagging, an image is expressed with words, while, in the case of captioning, a sentence can be created in connection with words. For this reason, it is possible to obtain image information in detail through captioning, compared to tagging. However, when a caption is created with words, there is the limitation of end-to-end to lower performance if labeled data are not sufficient. As a solution to the problem, meta-learning, in which a small amount of data can be used, is applied. This study proposes the captioning model based on meta-learning using prior-convergence knowledge for explainable images. The proposed method collects multimodal image data for predicting image information. From the collected data, the attributes representing object information and context information are used. After that, with the use of a small amount of data, meta-learning is applied in a bottom-up approach to creating a sentence for captioning. It can solve the problem caused by data shortage. Lastly, for the extraction of image features, LSTM for convolution network and captioning is established, and the basis for explanation is generated through the reverse operation. The generated basis is an image object. An appropriate explanation sentence is displayed in line with a particular object. Performance evaluation is conducted in two ways for accuracy. Firstly, BLEU score is evaluated according to whether there is meta-learning. Secondly, the proposed captioning model based on prior knowledge, RNN-based captioning model, and bidirectional RNN-based captioning model is evaluated in terms of BLEU score. Therefore, through the proposed method, LSTM’s bottom-up method reduces the cost of improving image resolution and solves the data shortage problem through meta-learning. In addition, it is possible to find the basis of image information using subtitles and to more accurately describe information about photos based on XAI.

Digital twins are virtual replicas of real physical entities in computers. They can be considered as abstract digital models of data and behavior for objects of interest. Nevertheless, they are not perfectly consistent with conventional data or simulation models because they achieve prediction and optimization by simulating the abstract digital model of a particular system. To maintain the characteristics of digital twins in the virtual space, digital simulation models that continue to update, change, and evolve according to continuous changes of corresponding physical factors must be used. Owing to the various advantages of digital twin technology, digital twins have gained more attention. However, the method to create digital twins is still unclear. Additionally, the availability and sufficiency of information on physical entities to which digital twins will be applied must be considered, and a model suitable for their application must be designed. Therefore, multimedia knowledge-based bridge health monitoring using digital twins is proposed herein. It synchronizes real and virtual spaces to reflect the reality based on various data collected using sensors of real systems. In this study, various situations of virtual bridge twins in a facility management area are simulated to provide digital services to ensure bridge health. This digital bridge health service analyzes situations based on a small amount of data collected from a bridge, predicts the optimal time point for maintenance, and then applies it to the real world. Hence, maintenance costs can be reduced and the bridge’s lifespan extended.

With the development of information technology, ambient intelligence has been combined with various application areas so as to create new convergence service industries. Through IT convergence, human-oriented technologies for improving people’s quality of life has continued to be developed. Healthcare service that has been provided along with the development of various smart IT devices makes it possible to realize more efficient healthcare of people. Therefore, along with such a medical service, the advanced lifecare service for physical and mental health has been demanded. In order to meet the healthcare demands, an advanced healthcare platform has been developed. Lifecare service has been expanded to healthcare, the disease with the highest mortality induced by complications so that the service for disease survivals have been offered. Accordingly, a big-data integration and advanced healthcare platform based on patients’ life logs are developed in order for health service. In this platform, it is possible to establish an optimized model with the knowledge base and predict diseases and complications and judge a degree of risk with the use of information filtering. The conventional filtering based on a data model using scatter life logs makes use of user attribute information only for clustering so that it has low accuracy. Also, in calculating the similarity of actual users, such a method does not apply social relationships. Therefore, this study proposes a social mining based cluster process for big-data integration. The proposed method uses conventional static model information and the information extracted from the social network in order to create reliable user modeling and applies a different level of weight depending on users’ relations. In the clustering process for disease survivals’ health conditions, it is possible to predict their health risk. Based on the risk and expectation of healthcare event occurrence, their health conditions can be improved. Lifecare forecasting model that uses social relation performs social sequence mining using PrefixSpan to complement the weak point that spends a long time to scan it repeatedly in the candidate pattern. For performance evaluation, the social mining based cluster process was compared with a conventional cluster method. More specifically, the estimation accuracy of the conventional model-based cluster method was compared with the accuracy of the social mining based cluster process. As a result, the proposed method in the mining-based healthcare platform had better performance than the conventional model-based cluster method.

For service and multimedia processing that are not limited by time and space, it is necessary to go beyond the existing computing paradigm and resolve such limitations. In this study, health big data-based cardiac disease induction prediction made with multimedia extraction is suggested, which analyzes the relationships in health big data using multimedia extraction. Multimedia extraction is roughly divided into two types: extraction of structured data–based significant items, and extraction of unstructured data–based information. The extraction of structured data–based significant items is made with a multivariate analysis algorithm and similarity analysis. The extraction of unstructured data–based information is made with a technique called parsing based on medical keywords. Using personal health record (PHR)-based data, health big data are collected, while items having significant relationships are selected using logistics regression. Depending on the proximity of the Minkowski distance, a risky group with high similarity to patients with cardiovascular diseases is formed, while risk factors for cardiovascular diseases are evaluated using the similarities between the risky group and the user. A multivariate analysis was used to analyze the items with a significant level of significance. Through this, 27 out of 210 items were extracted. Therefore, only 12.9% of the data are used, and with the MAE results, it was found that an error in accuracy of 0.21. These results show that the suggested model could provide more personalized data and can be used as core technology for constructing an effective, efficient, smart healthcare system.

In this paper, a method for knowledge expansion of metadata using script mining analysis for multimedia recommendation systems is proposed. The method allows the extraction of new metadata and knowledge expansion through the mining analysis of multimedia scripts, which include a large amount of information. The scripts are collected by a Web crawler based on Python. From the collected scripts, hidden information is extracted through keyword analysis and sentiment analysis. In keyword analysis, scripts, unlike general documents, show a high frequency of names of characters or proper nouns. Such names or proper nouns are not frequently used in other media content, and therefore, their importance is high. Frequently, they are already offered in the conventional metadata, and consequently cause information duplication. Accordingly, term frequency–inverse document and metadata frequency (TF–IDMF), which considers the frequency of metadata in general term frequency–inverse document frequency (TF–IDF), is used. Thus, the importance of the names of characters or proper nouns in scripts can be decreased. Because the keywords for the extracted scripts are in fact included in the scripts, they can be used for precise multimedia search and recommendation. In sentiment analysis, the AFINN lexicon and the Bing lexicon are utilized to scan words in a script. The Bing lexicon is used to examine whether the words in the entire script are positive or negative. Then, the total numbers of positive words and negative words are used to calculate the representative sentiment of the script. The AFINN lexicon includes approximately 170 sentiment words, the negative or positive sentiment of which is presented in the range − 5 to +5. One script is divided into 100 sentences, and then, the representative sentiment in each sentence is evaluated as either positive or negative. Through script scanning, the flow of sentiment in multimedia streams can be discovered. The Bing lexicon categorizes words into positive, negative, and neutral sentiments. Through script scanning, the words included in each category can be quantified. Depending on the result of the script sentiment analysis, a different sentence embedding method based on inter-sentence similarity is used to cluster similar media. The results of the keyword analysis and sentiment analysis of a script are added to the metadata in a new column in a knowledge base to expand knowledge. To evaluate the significance of multimedia recommendations, keywords and sentiment information are used, and then, the similarity and clustering of the extracted media are assessed. As a result, script mining analysis based on the attributes that include actual information of media is considerably better than that based on types or a range of metadata attributes. Therefore, the proposed knowledge expansion method achieves significant results and shows an excellent performance in multimedia recommendation.

This study proposes the multimedia recommendation method using Word2Vec-based social relationship mining. This is to analyze users with a similar tendency on the basis of the keywords related to multimedia content and sentiment words of comments, to build a trust relationship, and to recommend multimedia. In order to solve the problem of data sparsity, metadata of multimedia content are collected and then are clustered by genre. User’s evaluate a preference for multimedia content. With the use of evaluation data, the attributes preferred by users are predicted. In terms of propensities, the sentiment words in users comments are classified by SVM on the basis of sentiment dictionary. The classified sentiment words are presented in vector with the use of Word2Vec. In terms of the vector of sentiment words, the dynamic relationship between users of words in the same preference by the similarity using the distance scale. It helps to build a trust relationship between users with preferences that can change with a lapse of time. Accordingly, multimedia content are recommended to users with a similar tendency. In terms of performance evaluation, F-measure is compared with the uses of precision and recall for a recommendation. As the result of evaluation, the social relationship mining method is evaluated to be better than explicit and implicit recommendation methods. With the proposed method, it is possible to search with metadata of content and make a intelligent recommendation explicitly and implicitly according to user’s tendency.

Depression is a mental illness influenced by various factors, including stress in everyday life, physical activities, and physical diseases. It accompanies such symptoms as continuous depression, sleep disorder, and suicide attempts. In the healthcare, it is necessary to predict diverse situations accurately. Accordingly, in order to care for mental health, it is necessary to recognize individuals' situations and continue to manage them. In the area of mental diseases and treatment, research has been conducted to find a patient's state with the use of big data and to monitor the worst situation. Mental illnesses typically have depression. Research on Mental healthcare using artificial intelligence do conduct on prediction based on patients' voice, word choice, and conversation length. However, there is not much research on situation prediction in order to prevent depression. Therefore, this study proposes the context-DNN model for predicting depression risk using multiple-regression. The context of the proposed context-DNN consists of the information to predict situations and environments influencing depression in consideration of context information. Each context information related to predictor variables of depression becomes an input of DNN, and variable for depression prediction becomes an output of DNN. For DNN connection, the regression analysis to predict the risk of depression is used so as to predict the potential context influencing the risk of depression. According to the performance evaluation, the proposed model was evaluated to have the best performance in regression analysis and comparative analysis with DNN.

There is a rapidly growing interest in health care due to the recent development of IT convergence technologies according to the 4th industrial revolution. More services for personal health management of users are available and studies on the establishment of knowledge base for an efficient health log data management in the health care field are being carried out with the emergence of block chain technology which is the next generation information security technology. In this paper, a knowledge-based block chain network for health log data management mobile service is suggested. The user’s log data and context information are applied to block chain technology that is difficult to forge and falsify in the knowledge-based health platform, enabling a large amount of users’ log data and context information accumulated continuously to be stored in a block in the knowledge base using the side chain structure that stores information through the configuration of knowledge-based data transaction. This enables high expandability and security to be secured in mobile environment as well. The result of comparative evaluation with the existing ontology knowledge model for verifying the validity shows that the suggested method presented approximately 16.5% higher performance in accuracy and reproducibility.

In modern society, the number of chronic patients is increasing due to various causes, such as drinking, smoking, unhealthy lifestyles, and stress. Chronic diseases must be managed with constant care, but may get worse from various factors. With the development of information technology, healthcare technologies using health big data, machine learning, and reinforcement learning are attracting attention. Using these technologies, it is possible to predict potential diseases that may occur in the future by using data learning and clustering of similar data. To predict the potential for disease, we should research various models based on the convolutional neural network (CNN), which can identify knowledge objects from unstructured data such as medical data. However, the fully connected network structure of the CNN generally uses a large amount of memory. Another problem is that complexity increases with the number of layers. This causes the overfitting problem, which increases error. To solve this problem, this paper proposes a CNN-based health model using knowledge mining of influencing factors. The proposed method uses hidden layers of a double-layer structure within the CNN structure. The double-layer structure has the optimal conditions for classification, compared with a single layer that allows the AND/OR operations. First, the amount of data used is reduced by extracting influencing factors through multivariate analysis, and these influencing factors are used as input data. Significant influencing factors are extracted from the first hidden layer using the significance level. This improves accuracy, because it extracts data required for analysis. Common influencing factors appropriate for significance levels are extracted. Common influencing factors refer to correlated factors that can affect each other. In the second hidden layer, the correlations between influencing factors are discovered through a correlation coefficient, and they are classified into positive and negative factors. Furthermore, associated rules are discovered through knowledge mining from among the classified influencing factors. They are subdivided into influencing factors like obesity, high blood pressure, and diabetes through the rules of the discovered influencing factors. For performance evaluation, the root mean square error (RMSE) of the CNN model is evaluated according to the application of knowledge mining to the influencing factors. The evaluation of accuracy, computational load, complexity, and learning rate showed better results, compared with the existing method. Through the proposed health model, knowledge about the associations of various factors is derived.

BACKGROUND Due to environmental factors such as nutrient intake imbalance, lack of exercise, and increased stress, it is necessary to control nutrition in order to prevent diseases and provide treatment in terms of healthcare. OBJECTIVE This study proposes the activity-based nutrition management model with the use of the cluster analysis of similar group for healthcare. METHODS The proposed method is to conduct the cluster analysis of similar group for nutrition management and to develop the real-time activity information based nutrition management algorithm with the use of big data in order to improve the quality of healthcare management service. It is to re-process an existing nutrition database and add voice recognition function in line with the service so as to improve convenience of intake-food inputs. In addition, the Bluetooth Low Energy (BLE) communication based standard collection of bio signals occurring in real-time is developed. This study also proposes the method of improving an existing algorithm of drawing a daily recommended allowance with the use of real-time activity information, and the proposed service provides the essential information of nutrition management with the use of public big data. RESULTS To verify the developed technology and service model and its effectiveness, the nutrition management service system is designed and developed with human interface. CONCLUSIONS The developed health model helps to solve the obesity problem, save medical costs, and address the issue of national health.

Much attention has recently been given to changes in medical services, such as remote medical services and healthcare services customized for users, where cloud technology is utilized in the health and medical industries. It is possible to utilize health information needed by users in real time if medical records and health information are saved through this, so that the level of medical practice increases, and medical information in a new form can be predicted and provided based on big data analysis and processing through the accumulated medical data. Thus, studies of mobile services utilizing this are being conducted. In particular, there is an increasing demand for the development of a cloud-based service in pharmaceutical management. With the development of modern medicine, by simply taking medicine it has become possible to treat diseases that in the past might have threatened lives. When patients get prescriptions for various medicines, it is necessary to know what role and effect they have, and in order to prevent misuse and abuse, it is necessary to provide correct and accurate information about these medicines. Numerous drugs enhance quality of life, but those aimed at treatment may be lethal unless patients know how to take them. In this study, we propose a peer-to-peer (P2P)-based open health cloud for medicine management. The proposed system is designed to smoothly provide a virtual cloud service by building up various cloud environments and communicating with cloud servers in a pre-reserved, on-demand method. The aim is to resolve the problems of data processing and reduce delays with wireless body area networks (WBANs), which occur in existing cloud health services, and to enhance stability and QoS for things like response time. In addition, to integrate personal health record (PHR) data stored in an internal medical database built up by each institution, and for the integration of various medical systems, the proposed system is designed to allow access to file managers through open database connectors, and it includes source connectors that link to external systems. Based on this, an access interface was designed to provide a mobile chat-bot service for user convenience. The proposed system is a mobile health service in the form of a chat-bot to quickly deal with changes through incidents that may occur because users take the wrong medicine by mistake in their everyday lives. To build up and analyze big data based on information collected in various ways, and to provide the data to users conveniently, the service is broadly divided into five classes, and in each class and in each service class is a customized user interface/user experience.

Due to recent growing interest, the importance of preventive and efficient healthcare using big data scattered throughout various IoT devices is being emphasized in healthcare, as well in the IT field. The analysis of information in healthcare is mainly prediction using a user’s basic information and static data from a knowledge base. In this study, a knowledge-based dynamic cluster model using a convolutional neural network (CNN) is suggested for healthcare recommendations. The suggested method carries out a process to extend static data and a previous knowledge base from an ontology-based ambient-context knowledge base beyond knowledge-based healthcare management, which was the focus of previous study. It is possible to acquire and expand a large amount of high-quality information by reproducing inferred knowledge using a CNN, which is a deep-learning algorithm. A dynamic cluster model is developed, and the accuracy of the predictions is improved in order to enable recommendations on healthcare according to a user environment that changes over time and based on environmental factors as dynamic elements, rather than static elements. Also, the accuracy of the predictions is verified through a performance evaluation between the suggested method and the previous method to validate effectiveness, and an approximate 13% performance improvement was confirmed. Currently, the acquisition of knowledge from unstructured data is in its early stages. It is expected that symbolic knowledge-acquisition technology from unstructured information that is produced and that changes in real time, and the dynamic cluster model method suggested in this study, will become the core technologies that promote the development of healthcare management technology.

Because of the growth of the business sector dealing in the distribution of movies, software, music, and other contents, a very large amount of contents has accumulated. Accordingly, recommendation systems for inducing user requests for contents are more important. In distribution businesses, accurate content recommendations are required to secure and retain users. To establish a highly accurate recommendation system, the recommended contents must be accurately classified. As classification methods, mainly techniques such as naive Bayes, SGD (stochastic gradient descent), and SVM (support vector machine), are utilized. If all of the information on recommended subjects is applied in the classification process, high-level accuracy can be expected, but heavy calculation, a long service time, and low scalability are incurred. Given this inefficiency, effective classification in which the metadata of contents are used is required. Metadata are expressed in the forms of the domain concept, relation, type, and attribute to allow the complicated relations between multimodal data (text, images, and video) to be processed efficiently. Most classification systems use single modal data to express one piece of knowledge for an item in a domain. Single modal data are limited in terms of improving classification accuracy, because they do not include the useful information provided by different knowledge types. Therefore, in this paper, we propose MMCNet, a deep learning–based multimodal classification model that uses dynamic knowledge. The proposed method consists of a classification model that applies the human learning principle-based CNN (convolution neural network) to multimodal data in combination with text and image knowledge. By using a Web robot agent, multimodal data are collected from the TMDb (The Movie Database) data set, which includes a variety of single modal data. In the preprocessing procedures, knowledge integration, knowledge conversion, and knowledge reduction are performed to create a quantified knowledge base. To handle text data, sentences are refined through morphological analysis and converted to numerical vectors by using word embedding. Image data are converted to numerical vectors using the library related to vector conversion. The converted feature vectors are utilized to create multimodal learning data and the classification model is used for learning. To solve the problem of memory operation resources, vector model-based meta-knowledge is expanded through expression, conversion, alignment, inference, and deep learning. To evaluate its performance, the proposed model was compared with conventional classification methods in terms of accuracy, recall, and F1-score. According to this evaluation, the proposed classification model improves the accuracy, recall, and F1-score rates more than the conventional methods. In addition, the proposed model was implemented as a deep learning–based multimodal classification system in a graphical user interface environment that allows users to provide feedback about the classification results by adjusting classification parameters. Through the convergence of the knowledge bases of various domains and multimodal deep learning, the dynamic knowledge that influences user preference is inferred.

Grasping the sense of web pages on small screen of mobile devices is too difficult, so mobile users prefer summarized reports. One popular technique to generate automatic summarized report is based on vector model. The vector model summarization methods are simple and low cost to implement on mobile devices. However, the quality of summarized reports of the model may not be good because there is a semantic difference between manual summarized reports and the machine summarized reports. In addition, the censored documents of the centralized search engine cannot accurately reflect the query of user in the summary results. To overcome this constraints, automatic mobile device summarizer using a semantic feature of pseudo relevance feedback and P2P web search engine is proposed. The proposed method increases the quality of summarized reports by using the latent features of documents and clustering technique. The method uses none censored documents of the P2P web search engine to more accurately reflect user requirements in summarized reports. The automatic summarizer consists of a server-side containerized auto-summary module for flexible management of the summarization functions and a user-side mobile device module to reduce the overload of the mobile device.

With the development of ubiquitous computing, people easily access and share a variety of information through searches. Based on the social streams collected through the news, Twitter, web, internet communities, SNS (social network services), and internet boards, accurately searching for information in accordance with the user preferences is necessary. The volume of accumulated social streams rises rapidly with time, and the quality of the referred contents tends to be lowered by information noise, despite their frequencies. Therefore, this study focuses on the discovery of knowledge of associative relations using opinion mining on issues related to health. The proposed method mines rules and discovers knowledge through association analysis and opinion mining of social streams. Correspondingly, unstructured data for three major chronic diseases, namely, high blood pressure, diabetes, and hyperlipidemia, are collected with the use of a crawler. The extracted corpus is used to create transactions, and the association rules of the health corpus are mined. Sets of words with associations are organized to support the decision-making for the choice of words for use in a search engine. The mined association rules of the health corpus are based on relations of words, and meaningful relations are discovered based on opinion mining, i.e., based on a method of analyzing positive or negative aspects of formulated expressions in documents. To achieve this, vocabularies of a sentiment dictionary are used to calculate a frequency-based polarity value and a term frequency–inverse rule frequency (TF–IRF) weight. With the calculated polarity value and TF–IRF weight, the degree of opinion for specific words is drawn from association rules. In this manner, it is possible to express a positive or negative relation between words in a visual manner. Accordingly, the use of association rules and opinion degrees allows the generation of an opinion tree. This helps the conduct of an efficient information search for matters related to health and the formulation of opinion relations from an opinion knowledge tree so as to support decision-making. For performance evaluation, predictions were made in regard to the proposed method, and opinions of test sentences were evaluated. As a result, the precision and recall were excellent. By applying the opinion mining–based knowledge to matters relevant to health, it is possible to reach an accurate decision. In addition, with an inference engine, it is possible to provide a customized UI/UX in an ambient context and thus create added value in health services.

Big-data knowledge processing technology facilitates efficient health management services by systematically collecting and promoting information using distributed/parallel processing with the health platform’s common data model. Thus, it enables knowledge expansion for healthcare data. In this study, we propose a big-data knowledge process for the health industry using Hadoop’s MapReduce software for association mining. The proposed method provides efficient health management knowledge services by collecting and processing heterogeneous health information using WebBot and the common data model. Hadoop is a proprietary method of effectively processing distributed big data. It is a knowledge processing model that combines MapReduce-based distributed processing and a method of finding mining-based associations. The input data in MapReduce is extracted from chronic disease nomenclature from health big data. The corpus divides big data into several blocks of a certain size, creating map tasks. Through the map function of the mapper of each map task, sets composed of pairs of a key and a value are created. In the map process, a key is created using the same method used for a frequent item set of the Apriori algorithm. The key is a set of 2p keys and its value is set to the occurrence frequency of the key. By summing up the values of the same keys by combining, the size of data is decreased and the load of a software program is also decreased. In addition, for each key, the reducer is designated through hash partitioning and stored in the reduce task. In the reduce process, the results of the map are allocated to each reducer, and alignment and merge steps are taken based on the keys. For the same |key|, the values are summed up by performing the reduce function. In this instance, keys whose values fail to meet the minimum support criterion are eliminated. Therefore, from a set of , a frequent item set that meets the minimum support criterion is extracted. The association rules between datasets constituting the frequent item set are determined, and the support and reliability are calculated to examine whether they are actually associated. As the value of the frequent item set is higher, the support and reliability are also higher. Thus means that the association is obvious. A knowledge base is then constructed using the extracted association rules by repeatedly performing the MapReduce process. Closely associated knowledge bases are created and semantically related in real time with high probability. Furthermore, mining-based knowledge processing of health big data infers more meaningful associations between chronic diseases. The proposed method adds technological value and intelligent efficiency to support the health and medical fields promote healthy lives.

With the change in their social environment and life patterns, the eating habits of modern people have become more diverse. Eating habits are closely related to health, and diet management for individual healthcare is required in this era characterized by poor health and increased longevity. In this paper, we propose a knowledge-based hybrid decision model for nutrition management that uses neural networks. The proposed method is a food recommendation model to help users make dietary nutrition-related decisions on a health platform. It is a hybrid recommendation method that considers both physical and mental health. It selects foods that are positively related to users’ physical health as candidates and predicts users’ preferences through adaptive learning. A previously developed dietary nutrition service ontology is used to select foods that appear to affect the user’s health positively. Conventional preference prediction methods include collaborative, content-based, knowledge-based, and image-based filtering. These methods use a hybrid model or machine learning, data mining, and artificial intelligence methods to compensate for the disadvantages of each filtering type. For preference prediction, healthcare and food preference data are collected in an on/off line environment. The data consist of age, sex, body mass index, region, chronic disease, and food preferences. Food preferences include the dietary nutritional components of food, which makes it possible to infer the user’s preferences for foods containing calories, carbohydrates, protein, fat, sugars, sodium, cholesterol, saturated fatty acids, and trans fatty acids. The user’s preference for food is composed of output variables, and other variables are composed of input variables. The variables consist of 11 healthcare data variables, 2 preference data variables, 10 dietary nutrition data variables, 22 input variables, and 1 output variable. The variables that we constructed are used to arrange transactions and supervised learning is conducted in a neural network structure. In total, 3152 transactions, 80% of the collected data, were used as learning data and 788 transactions, 20% of the collected data, as test data. Using the test data, we evaluated the performance of four prediction models based on a learned neural network, user correlation, average replacement, and regression analysis, respectively. The result of the performance evaluation showed that the proposed method was superior to the conventional method in that it solved the cold-start and the sparsity problem. In addition, the user’s satisfaction evaluation result was 3.92 on a five-point scale, showing overall satisfaction. Therefore, on the platform it is possible to recommend dietary nutrition for people suffering chronic diseases according to their lifestyle and in consideration of their health status and preferences. The platform selects a suitable candidate food according to the health condition of the user and provides a recommendation for N foods using the Top-N of the user’s food preferences.

Currently, significant amounts of IoT data are being collected as big data for deep neural network learning algorithms that are used to extract meaningful information from big data and applied to various areas. However, a neural network’s over-fitting problem and rising computational costs associated with high levels of time complexity are obstacles to deep learning computations. Because of these problems, response delays are prevalent in big data learning processes and data extraction processes that use deep neural networks, which exponentially increase the cost of data extraction. Therefore, the amount of computation must be controlled so that data processing models can be used practically, and various dedicated devices can be used to process big data, including methods such as distributed processing methods. In general, a P2P method is a typical distributed processing method. P2P is based on traffic distribution and cooperation, and edge computing is a method that increases efficiency by locally processing large volumes of data produced and only transmitting essential information. If P2P method-based edge computing techniques are used, an effective parallel processing model can be constructed, reducing the computing requirements of a central server and the load of a network. In this paper, we propose an edge computing health model using P2P-based deep neural networks. The proposed method is used to process health big data in edge computing. To construct the model, multiple edge nodes are required, and the edge node modules are deployed at locations where health data are produced and directly connected to the deep neural network model. At the edge nodes, a modularized deep neural network is constructed, thereby implementing a parallel big data processing system. In addition, a single server is required for gathering the results of a neural network model, and because the server only collects evaluation results and provides them to the users, the response time delay is improved. In this paper, to evaluate an improved response time, a regular server model and an edge computing health model are constructed separately, and positive numerical values are extracted from the experiments. The extraction results confirm that a combination of parallel processing models and deep neural network techniques can be used to distribute and process computing operations of big data size, ensuring an effective system for reducing response time delay.

Lifecare for quality-of-life improvement includes not only the prevention and treatment of diseases and health management anywhere and at any-time, but also the improvement of lifestyle, the psychological environment, and the emergency response. The construction of a sustainable model on the telecommunications platform as a solution has been constantly studied using intelligence information technology. In this study, sequential-index pattern mining on the lifecare telecommunications platform is proposed. The method is as follows. Various temporal and regional indices that are provided by the Meteorological Administration, Meteorological Data Open Portal, Public Data Portal, Healthcare Big Data Hub, and Korea Environment Corporation are collected as lifecare index data. The collected index data are transmitted to the NAS file server of the lifecare telecommunications platform using peer-to-peer (P2P)-based interconnection technology and high-performance computing. The sequential relationships among the indices are discovered to consider the semantic relationships. Using the AprioriAll mining algorithm, the maximum sequence is determined to find sequential relationships in the frequent sequence set. To process the pattern extraction and computation efficiently, recomposed transactions are stored in the high-performance NAS file server. Also, a mining-index prediction model is developed in consideration of the sequential relationships to complement the weakness of the statistical time-series analysis. Since the regional meteorological conditions are similar, the index sequential pattern can solve the problem of the generation of omissions or errors regarding the real-time monitoring data that are due to the monitoring equipment or communication failures. Therefore, the proposed sequential-index pattern mining can be utilized as a lifecare tool for risk-factor detection, risk prediction, and trend analysis.

Cognitive manufacturing has brought about an innovative change to the 4th industrial revolution based technology in combination with blockchain distributed ledger, which guarantees reliability, safety, and security, and mining-based intelligence information technology. In addition, artificial intelligence, machine learning, and deep learning technologies are combined in processes for logistics, equipment, distribution, manufacturing, and quality management, so that an optimized intelligent manufacturing system is developed. This study proposes a topic mining process in blockchain-network-based cognitive manufacturing. The proposed method exploits the highly universal Fourier transform algorithm in order to analyze the context information of equipment and human body motion based on a variety of sensor input information in the cognitive manufacturing process. An accelerometer is used to analyze the movement of a worker in the manufacturing process and to measure the state energy of work, movement, rest, and others. Time is split in a certain unit and then a frequency domain is analyzed in real time. For the vulnerable security of smart devices, a side-chain-based distributed consensus blockchain network is utilized. If an event occurs, it is processed according to rules and the blocking of a transaction is saved in a distributed database. In the blockchain network, latent Dirichlet allocation (LDA) based topic encapsulation is used for the mining process. The improved blockchain distributed ledger is applied to the manufacturing process to distribute the ledger of information in a peer-to-peer blockchain network in order to jointly record and manage the information. Further, topic encapsulation, a formatted statistical inference method to analyze a semantic environment, is designed. Through data mining, the time-series-based sequential pattern continuously appearing in the manufacturing process and the correlations between items in the process are found. In the cognitive manufacturing, an equalization-based LDA method is used for associate-clustering the items with high frequency. In the blockchain network, a meaningful item in the manufacturing step is extracted as a representative topic. In a cognitive manufacturing process, through data mining, potential information is extracted and hidden rules are found. Accordingly, in the cognitive manufacturing process, the mining process makes decision-making possible, especially advanced decision-making, such as potential risk, quality prediction, trend prediction, production monitoring, fault diagnosis, and data distortion.

Humans require rest in order to recover from physical and psychological fatigue caused by daily routines, and either medication or sleep is needed to achieve this. If there is a lack of sleep, many health issues can occur, such as depression, anxiety, obsession, dementia, deteriorated immunity, and lowered physical ability. In modern society, a variety of sleep disorders including insomnia, hypnolepsy, sleep apnea, and anxietas tibiarum, occur depending on the environment or occupations. As such, sleep significantly influences the quality of life, and it is a fundamental factor of healthcare. Thus, to increase the quality of sleep, research has been conducted on sleep disorder treatment in various fields including humanities, social studies, science, and engineering. In addition, data collection and pattern analysis of sleep have been researched and developed in detail. Although diverse sleep management products using IT convergence technology have been released, they only provide simple numbers or graphs of the personal sleep data measured. Therefore, this study proposes a mining-based time-series sleep pattern analysis of life big data. The proposed method collects and pre-processes sleep logs obtained continuously from life big data over time. As a result of pre-processing, sleep indexes comprise total sleep time, light sleep, deep sleep, awakening time, and bed time. External and internal factors that influence a sleep pattern are designed as context information, and then are normalized. In addition, context information is processed as a context index to create sleep data transactions. Based on the context index, similar users are identified based on similar context information. An expectation maximization algorithm is used to calculate maximum a posteriori through repetition of the expectation step, and a maximization step is applied for clustering. In sleep data transactions of a user set of a cluster, the AprioriAll algorithm of data mining is applied to create a sequential pattern of sleep in order to find meaningful associations, patterns, and trends. A pattern fitting a sleep stage is analyzed in order to improve sleep habits and treat sleep disorders. To improve REM sleep and ensure that a sufficient number of sleep hours are obtained, a mining-based time-series pattern is offered as feedback to each user. In this way, it is possible to improve the quality of sleep and life of a user.

Context computing is a branch of ambient intelligence (AmI) research, which has been rapidly emerging in the support of intelligent smart health platform solution. To develop reliable ambient computing using the hybrid peer-to-peer network and Internet of Things, machine learning, deep learning, artificial intelligence, and context awareness have been applied. This study proposes an ambient context-based modeling for a health risk assessment using deep neural network. In the proposed method, we collected medical information from chronic disease patients such as EMR, PHR, and medical histories, as well as environmental data from a health platform. Subsequently, heterogeneous data are integrated through selecting, cleaning, modeling, and evaluating the collected raw data and then the context is created. The structured input data such as a sensor data are normalized by transforming the time domain data to the frequency domain information. Using a deep neural network, the normalized data are applied to create an ambient context. A deep neural network is composed of the following three layers: input layers with treated and untreated data; hidden layers where connection strength is trained as a weight; and output layers of trained results. In the deep neural network layers, the control of the weight of training data enables repeated learning to create an ambient context pattern. Using an ontology inference engine, unstructured/structured data, including individual health data and environmental information, and their context is presented as ontology metadata. In the knowledge base, hidden association relationships are discovered through mining. To inform the individual health conditions exposed to the individual environmental contexts, a health risk assessment model is developed with a set of the ambient context pattern learned with metadata and a deep neural network. The Minkowski distance formula, which defines a normalized geometrical distance between two nodes, is used to measure the similarity between the patients with chronical disease and the individual user based on the context. In the proposed model, the risk is represented as a similarity-based index. The risk assessment model can be implemented into the individual risk alert/prevention system. The model may significantly impact the healthcare industry as well as ambient intelligence research, thus contributing to improve the quality of human life of the future society.

To overcome the limitations of the conventional medical service in terms of ageing and chronic diseases, AmI-based precision medicine has drawn particular attention. Precision medicine is a customized personal medical service using various information technologies such as personal health device, AI algorithm, image recognition, voice recognition, and natural language processing. In particular, the information technologies for follow-up care services for patients, such as context awareness, context information, and inference rules, are required. In PHD, contexts such as variable data include blood pressure, BMI, blood sugar, weather, and food. It has time-series characteristics, meaning that it changes often with time. Other kinds of health-related information, such as age, family history, smoking, and residential area, are intermittently changed. For inference that is highly related to a user, the context collected through AmI is presented with ontology. Ontology consists of a user’s ambient data, weather data, and lifelog. Context is changed along with a user’s ambient conditions and time. An inference engine is used to create the knowledge base and predict a change. This study proposes a neural-network based adaptive context prediction model for ambient intelligence. This is a learning model using neural network to calculate the similarity for recommendation in a mining lifecare platform. In a conventional prediction procedure, an error is used to update a weight. The proposed model learns the similarity weight of the users to become adapted to the user’s ambient. Based on the knowledge base, user clustering and deviation from mean are applied to calculate the similarity weight. Collaborative filtering technology is used to predict a user’s context and learn the similarity weight repeatedly using a neural network. According to the performance evaluation, the proposed neural-network based similarity weight method had the highest accuracy of prediction when the learning rate was 0.001. Consequently, we found that AmI is a new added-value technology to maintain a healthy lifestyle and contributes to developing the healthcare industry and improving the quality of life.

With the development of big data computing technology, most documents in various areas, including politics, economics, society, culture, life, and public health, have been digitalized. The structure of conventional documents differs according to their authors or the organization that generated them. Therefore, policies and studies related to their efficient digitalization and use exist. Text mining is the technology used to classify, cluster, extract, search, and analyze data to find patterns or features in a set of unstructured or structured documents written in natural language. In this paper, a method for extracting associative feature information using text mining from health big data is proposed. Using health documents as raw data, health big data are created by means of the Web. The useful information contained in health documents is extracted through text mining. Health documents as raw data are collected through Web scraping and then saved in a file server. The collected raw data of health documents are sentence type, and thus morphological analysis is applied to create a corpus. The file server executes stop word removal, tagging, and the analysis of polysemous words in a preprocessing procedure to create a candidate corpus. TF-C-IDF is applied to the candidate corpus to evaluate the importance of words in a set of documents. The words classified as of high importance by TF-C-IDF are included in a set of keywords, and the transactions of each document are created. Using an Apriori mining algorithm, the association rules of keywords in the created transaction are analyzed and associative keywords are generated. TF-C-IDF weights and associative keywords are extracted from health big data as associative features. The proposed method is a base technology for creating added value in the healthcare industry in the era of the 4th industrial revolution. Its evaluation in terms of F-measure and efficiency showed its performance to be high. The method is expected to contribute to healthcare big data management and information search.

With the recent increase in the interest of individuals in health, lifecare, and disease, hospital medical services have been shifting from a treatment focus to prevention and health management. The medical industry is creating additional services for health- and life-promotion programs. This change represents a medical-service paradigm shift due to the prolonged life expectancy, aging, lifestyle changes, and income increases, and consequently, the concept of the smart health service has emerged as a major issue. Due to smart health, the existing health-promotion medical services that typically have been operated by large hospitals have been developing into remote medical-treatment services where personal health records are used in small hospitals; moreover, a further expansion has been occurring in the direction of u-Healthcare in which health conditions are continuously monitored in the everyday lives of the users. However, as the amount of data is increasing and the medical-data complexity is intensifying, the limitations of the previous approaches are increasingly problematic; furthermore, since even the same disease can show different symptoms depending on the personal health conditions, lifestyle, and genome information, universal healthcare is not effective for some patients, and it can even generate severe side effects. Thus, research on the AI-based healthcare that is in the form of mining-based smart health, which is a convergence technology of the 4IR, is actively being carried out. Particularly, the introduction of various smart medical equipment for which healthcare big data and a running machine have been combined and the expansion of the distribution of smartphone wearable devices have led to innovations such as personalized diagnostic and treatment services and chronic-disease management and prevention services. In addition, various already launched applications allow users to check their own health conditions and receive the corresponding feedback in real time. Based on these innovations, the preparation of a way to determine a user’s current health conditions, and to respond properly through contextual feedback in the case of unsound health conditions, is underway. However, since the previously made healthcare-related applications need to be linked to a wearable device, and they provide medical feedback to users based solely on specific biometric data, inaccurate information can be provided. In addition, the user interfaces of some healthcare applications are very complicated, causing user inconvenience regarding the attainment of desired information. Therefore, we propose a chatbot-based healthcare service with a knowledge base for cloud computing. The proposed method is a mobile health service in the form of a chatbot for the provision of fast treatment in response to accidents that may occur in everyday life, and also in response to changes of the conditions of patients with chronic diseases. A chatbot is an intelligent conversation platform that interacts with users via a chatting interface, and since its use can be facilitated by linkages with the major social network service messengers, general users can easily access and receive various health services. The proposed framework enables a smooth human–robot interaction that supports the efficient implementation of the chatbot healthcare service. The design of the framework comprises the following four levels: data level, information level, knowledge level, and service level.

In an era of many diseases and increased longevity, more attention has been paid to chronic diseases that require constant health care. Under this circumstance, the development of research and development (R&D) for smart-device-based constant health care has drawn great attention. With the emergence of wearable devices, personal health devices (PHDs), and smartphones, various contents for constant health care have been developed. By using these devices, the users are able to collect personal health records (PHRs) that include data such as activity amount, heart rate, stress, and blood sugar. The range of the collected PHRs can be limited depending on the equipment or the surrounding environment. To overcome this problem, it is necessary to make a comparison with similar users in a cluster. Also, it is necessary to provide a service that can analyze and visually display the collected personal-health information. In this paper, we propose the mining of health-risk factors using the PHR similarity in a hybrid P2P network. This is a method of predicting a user’s health status using similarity-based data mining, where the PHRs are employed in a hybrid P2P environment consisting of a peer, a server, and a gateway. In a hybrid P2P environment, a user receives feedback on the result of a structured-data analysis. A peer searches for a different peer and gateway through a server and exchanges information. Depending on the data type, the PHR is divided into medical health examination, self-diagnosis, and personal-health data. The medical health examination contains the personal-health data that are generated regularly by a medical institution. Self-diagnosis represents the data of mental health, pains, and fatigue that can be changed often but cannot be collected by devices. Personal-health data mean the data that can be collected by individuals in everyday life. For the PHR-data analysis, an index is given to each attribute, and preprocessing is performed after a binary-code conversion. To predict a user’s health status, the PHR data are clustered on the basis of similarity in a hybrid P2P environment. The similarity between a user’s PHR and a PHR that is searched for in the network is measured. After the measurement, an index is given to the PHR that meets the minimum similarity and the PHR is incorporated into a Similarity PHR Group. The Similarity PHR Group flexibly changes depending on a user’s PHR status and the statuses of the users who have accessed the hybrid P2P network. A representative value of the Similarity PHR Group is extracted and is then compared with the user’s PHR to judge the user’s health status. The proposed method is suitable for a smart health service for chronic diseases requiring constant care, elderly health, and aftercare. This is a user-oriented health-care and promotion service wherein a user’s health status can be predicted through the mining of the health-risk factors of PHRs.

Due to the enhancing of life quality, increasing of chronic diseases, changing lifestyles, and an expanding life expectancy, rapid population aging requires a new business model that promotes happiness and emphasizes a healthy body and mind through the “anytime, anywhere well-being” lifestyle. Recently, lifecare systems using IoT devices are being released as products that are influential on the overall society, and their effectiveness is continuously proven. In addition, based on peer-to-peer (P2P) networking, diverse companies are conducting investments and research to develop devices as well as solutions that connect to these devices. Accordingly, in this study, a mining-based lifecare-recommendation method using a peer-to-peer dataset and adaptive decision feedback is proposed. In addition to collecting PHRs, the proposed method measures life-logs such as dietary life, life pattern, sleep pattern, life behavior, and job career; the P2P-dataset preprocessed index information; and biometric information using a wearable device. It uses the Open API to collect the health-weather and life-weather index data from public data, and it uses a smart-band-type wearable device known as a biosensor to measure the heart rate, daily activity, and body temperature. It monitors the current status and conditions through the classification of life data, and it mines big data and uses a decision tree to analyze the association rules and correlations, as well as to discover new knowledge patterns. In the peer-to-peer networking, a lifecare recommendation model that uses adaptive decision feedback has been developed for the peer-to-peer platform. This adaptive decision feedback reflects an individual’s importance or sensory level. Accordingly, it proposes more individualized and flexible results and can be configured to support intellectual lifecare. A mining-based lifecare-recommendation mobile service can also be developed to enhance the quality of life, as it provides user-based health management and reduces the medical expenses; accordingly, it enhances the service satisfaction and quality in the lifecare field.

Today’s medical industry can be represented by a human-centered u-healthcare paradigm that is available and accessible anywhere, where high-tech IT can serve as the basis at any time and any place. In addition, in the medical industry, studies of many of the developments and applications are actively conducted based on the development of information communication technology. The aim of medical-information systems is the construction of an advanced IT and integrated u-healthcare system that evolves in the direction of integrated medical-based IT-convergence systems. Accordingly, to resolve the problems of telemedicine, in terms of the remote access to medical data, some public and private initiatives have been proposed, ranging from patient-mobility approaches to medical data. In addition, regarding GENICloud, which provides links with the existing future Internet testbeds and the Eucalyptus Cloud, two out of the seven GENIAM APIs have been announced as the common APIs of the future Internet testbeds in GENI, and they have been implemented and are provided. In the present GENI Cloud system, due to the provision of limited APIs, restrictions may occur in the future Internet testbeds and the Eucalyptus Cloud system management. Therefore, this study proposes a cloud-based mobile health service for the enhancement of the quality of service (QoS) including factors such as reliability and response time to resolve the problems of the broadband-communication infrastructure in the existing mobile health service and the delay problem on the wireless body area network. In this paper, we propose the cloud-based u-healthcare network with a QoS-guaranteed mobile health service. For this method, the TMO-distribution object model that was used in the existing research to implement a reliable and efficient cloud system for users was not used, and instead, a cloud-platform environment was built up through the construction of a distributed system based on a cluster-based mobile object. For this purpose, this study considered the characteristics of the wireless-communication environments between the terminals and the cloud servers in the mobile cloud environment and the proposed cloud mobility services and the specialized mobile cloud-control software. Later, for linkages with cloud computing environments and testbeds was proposed. In addition, this study carried out a cloud mobility-control design to provide a service in the mobile cloud environment that is based on the actual future Internet testbeds. Lastly, based on the structured cloud-platform environment, this study designed access interfaces to provide a mobile healthcare service in consideration of the user convenience. For the mobile-service access interfaces, since the same service interfaces can be used to access the characteristics and functions of all of the applications from browsers and device clients, the model-view-controller structure of the platform was designed, including the components for the further improvement of the requirements, reuse, and maintenance of the codes in medium and large distributed systems.

Although mammography is an effective screening method for early detection of breast cancer, it is also difficult for experts to use since it requires a high level of sensitivity and expertise. A computer-aided detection system was introduced to improve the detection accuracy of breast cancer in mammography, which is difficult to read. In addition, research to find lesions in mammography images using artificial intelligence has been actively conducted in recent days. However, the images generally used for breast cancer diagnosis are high-resolution and thus require high-spec equipment and a significant amount of time and money to learn and recognize the images and process calculations. This can lower the accuracy of the diagnosis since it depends on the performance of the equipment. To solve this problem, this paper will propose a health risk detection and classification model using multi-model-based image channel expansion and visual pattern shaping. The proposed method expands the channels of breast ultrasound images and detects tumors quickly and accurately through the YOLO model. In order to reduce the amount of computation to enable rapid diagnosis of the detected tumors, the model reduces the dimensions of the data by normalizing the visual information and use them as an input for the RNN model to diagnose breast cancer. When the channels were expanded through the proposed brightness smoothing and visual pattern shaping, the accuracy was the highest at 94.9%. Based on the images generated, the study evaluated the breast cancer diagnosis performance. The results showed that the accuracy of the proposed model was 97.3%, CRNN 95.2%, VGG 93.6%, AlexNet 62.9%, and GoogleNet 75.3%, confirming that the proposed model had the best performance.

Speech enhancement algorithms play an important role in speech signal processing. Over the past several decades, many algorithms have been studied for speech enhancement. A speech enhancement algorithm uses a noise removal method and a statistical model filter to analyze the speech signal in the frequency domain. Spectral subtraction and Wiener filters have been used as representative algorithms. These algorithms have excellent speech enhancement performance, but suffer from deterioration in performance due to specific noise or low signal-to-noise ratio (SNR) environments. In addition, according to estimations of erroneous noise, a noise existing in a voice signal is maintained so that a spectrum corresponding to a voice signal is distorted, or a frame corresponding to a voice signal cannot be retrieved, and voice recognition performance deteriorates. The problem of deterioration in speech recognition performance arises from the difference between speech recognition and training model. We use silence-feature normalization model as a methodology to improve the recognition rate resulting from the difference in the noisy environments. Conventional silence-feature normalization has a problem in that the silent part of the energy increases, which affects recognition performance due to unclear boundaries categorizing the voice. In this study, we use the cepstrum feature of the noise signals in the silence-feature normalization model to improve the performance of silence-feature normalization in a signal with a low SNR by setting a reference value for voiced and unvoiced classification. As a result of recognition rate confirmation, the recognition rates improve in performance, compared with other methods.

With the development of convergence information technology, all of the spaces and objects of human living have become digitized. In the health- and medical-service areas, IT supports Internet of things (IoT)-based medical services and health-care systems for patients. Medical facilities have been advanced on the basis of such IoT devices, and the digitized information on human behaviors and health makes the delivery of efficient and convenient health care possible. Under the given circumstances, health and medical care have been researched. For some of this research, the patient-health data were collected using IoT-based medical devices, and they served as a tool for medical diagnosis and treatment. This study proposes the development of a medical big-data mining process for which topic modeling is employed. The proposed method uses the big data that are offered by the open system of the health- and medical-services big data from the Health Insurance Review and Assessment Service, and their application follows the guidelines of the knowledge discovery in big-data process for data mining and topic modeling. For the medical data regarding the topic modeling, the public structured health- and medical-services big data, Open API, and patient datasets were used. For the document classification in the semantic situation of a topic, the Bag of Words technique and the latent Dirichlet allocation method were applied to find the document association for the development of the medical big-data mining process. In addition, this study conducted a performance evaluation of the topic-modeling accuracy based on the medical big-data mining process and the topic-modeling efficiency, and the effectiveness of the proposed method was examined.

As infinite competition and materialism have become severe in the current society, stress management has emerged as a main topic. There are many causes that create stress, including external factors and personal events. Also, stress has different levels, depending on an individuals’ subjective analysis. Stress has high correlations with cardiovascular disorders and mental illness. In particular, long-term stress leads to lowered immunity, which makes people more exposed to various diseases, and brings personal and social costs. With the rapid development of the IoT, it has been easy to analyze and manage stress with the use of sensors and communications technology relating to the human body and its surroundings. This study proposes a heart-rate variability-based stress index service using a biosensor. The proposed method collects a variety of information in dual physical environments (such as temperature, humidity, and brightness) from IoT devices, and analyzes it in real-time. The discomfort index and wind chill temperature index offered by the Korea Meteorological Administration, and the temperature, humidity, noise, and brightness collected from a biosensor are the most clear factors to digitize the physical environments of stress. Also, a smart health platform analyzes different heart rates depending on individual conditions, and monitors current status. For a heart rate, the frequency of the R-R value and low frequency (LF) are analyzed. For R-R value, a maximum value detection algorithm is applied. For LF analysis, Fourier transform is used. Generally, fast Fourier transform is unable to analyze the relation between time and frequency. Accordingly, applied is a short time Fourier transform in which window size is limited in a graph so as to express an effect made by changing time effectively. A stress index is comprised of discomfort level, wind chill temperature, noise, brightness, and heart rate. The notification of risk is given to the user by signal lights indicating stability, warning, or danger. The stress index service enables a user to check the stress index in real-time over a smart health platform at any place and at any time. Therefore, it serves as a tool to notify one’s acquaintances of a risk when one faces an emergent situation or is about to be at risk.

The development of the IT convergence technology has been serving as the basis for the application of the IoT devices to the products being used in our everyday life, such as clock, health band, scale, TV, light and doorlock. The government and companies have been developing the IoT-based smart home, smart city and smart health to enhance the national health as well as the life convenience. In particular, the government has been providing its diverse supports to the smart health in order to utilize electronic medical records, personal health record and self-diagnosis. The Ministry of Health and Welfare has been expanding the self-diagnosis service through conducting an epidemiologic survey on mental illness and mental health, the two issues that are becoming a major problem in our society. The results of the survey-based self-diagnosis may vary depending on the environment surrounding the users as well as the mental state of the users. In this paper, we propose the depression index service using the knowledge-based crowdsourcing within the smart health platform. The proposed method provides the index suitable for the context of each user through using the knowledge-based crowdsourcing within the smart health platform to compare the actual users with the similar users and predict the depression level. The prediction of the depression index uses the collaborative filtering model based on the comparison of users with other users, and uses the crowdsourcing to supplement the sparsity problem. CES-D is used as the criterion for determining the depression level, and the CES-D data is collected through the crowdsourcing. In the preexisting crowdsourcing process, the highly expert/reliable knowledge base to which the context information of the users is applied is constructed. The context information of the users is additionally collected through the self-diagnosis process, and the data are determined through the preconditioning process. The actual users data and determined data are applied to the Pearson's correlation coefficient to analyze the similarity level of between the actual/preexisting users. Then the analyzed similarity is used to service the result of the self-diagnosis to the actual users. This is a flexible service to which the context information of the users is applied, and it contributes to the users' decision-making.

Accessibility to the information increased by the development of IT converging technology and therefore various information such as economy, society, technology, culture and etc are distributed without limit to time and place. Data mining that discovers useful information by analyzing the data that accumulated to a large amount is receiving spotlight. By the distribution of IoT machine, the collection of life log has been eased and the interest on personalized service fit to the used is increasing. Companies and nationals use various analysis methods to develop big data that is accumulated for the safety of its people and asset into simple index. Meteorological Administration analyze the weather data and provide service through internet and app by developing weather index analyzing the extent of influence of weather on life, health and industry. The weather index measure by each region by using data collected real time at the weather observation point. Weather index is classified into the life weather index, health weather index and industry weather index by fields and provide coping measures per steps along with precautions. The weather index provided at this present is a simple service that did not consider the correlation of the index. Also it has a problem of not considering the situation by the location of the user. This study suggests emerging risk forecast system using associative index mining analysis. The suggesting method is the monitoring system focused on the user by analyzing the associative index using data mining and forecasting emerging risk by reflecting the user situation information. Index data collects the weather index XML file by using the Rest call method through the Meteorological Administration. The collected weather index analyzed the association relationship and interrelationship between the index by using the Apriori algorithm and Pearson correlation coefficient of the data mining. User situation information collects the surrounding temperature and humidity using the Zigbee communication. Collected temperature and humidity is applied to the daily weather index and health weather index and calculated fit for the situation of the user and the industry weather index is calculated focused on the user by using the association relationship and interrelationship between the index. By using the calculated weather index focused on the user, emerging risk of the life, health and industry scope is forecasted. Emerging risk is classified into life emerging risk, health emerging risk, industry emerging risk by the forecast method and is divided into four steps of low, general, high and danger. Emerging risk forecast system using the associative index mining analysis suggested provides precaution followed in the emerging risk and provides service such as weather index, weather information, user situation information etc of the Meteorological Administration. The suggested emerging risk forecasting system is the user based monitoring simulation that calculates the weather index real-time.

The modern society has been developing new paradigms in diverse fields through IT convergence based on information technique development. In the field of construction/transportation, such IT convergence has been attracting attention as a new generation technology for disaster prevention and management. Researches on disaster prevention and management are continuously being performed. However, the development of safety technology and simulation for prediction and prevention is comparatively slow. For the new generation IT convergence to efficiently secure safety and manage disaster prevention, it is more important than anything else to construct systematic disaster prevention system and information technology. In this study, we suggested the associative context mining for ontology-driven hidden knowledge discovery. Such method reasons potential new knowledge information through the association rule mining in the ontology-driven context modeling, a preexisting research, and uses the semantic reasoning engine to create and apply rules to the context simulation. The ontology knowledge base consists of internal, external, and service context information such as user profile, weather index, industry index, location information, environment information, and comprehensive disaster situation. Apriori mining algorithm of the association rule is applied to reason the potential relationship among internal, external, and service context information and discovers and applies hidden knowledge to the semantic reasoning engine. The accuracy and validity are verified through evaluating the performance of the developed ontology-driven associative context simulation. Such developed simulation is expected contribute to enhancing public safety and quality of life through determining potential risk involved in disaster prevention and quick response.

BACKGROUND Humans supply a variety of nutrients to their body in dietary life, which are directly related to health. Chronic diseases are long accumulated in the body on account of heredity or living habits, and draw attention as a main issue in the era of disease-controlled longevity. Therefore, it is essential to make health care continuously through the improvement in dietary habits. OBJECTIVE By recommending alternative food products whose diet and nutrition structure is similar to that of the food products positively influencing users' health conditions, it is possible to satisfy user's health and preference. METHOD We used the hybrid clustering based food recommendation method that uses chronic disease based clustering, diet and nutrition ontology, diet and nutrition knowledge base. Active users are classified into the chronic disease based cluster that has the nearest euclidean distance. According to the classified clusters, food products are recommended to users, and similar food products are also recommended with the use of food clustering and knowledge base. Food products are clustered with the uses of k-means algorithm and food and nutrient data system. Based on the created food clusters and food preference data, diet and nutrition knowledge base is generated. It is composed of food cluster filter, food similarity filter, universal preference filter, and user feedback filter. The universal preference filter represents the similarity weight between diet and nutrition, and user preference. The user feedback filter has the similarity weight between active user preference and diet and nutrition. They continue to be updated through associated feedback. RESULT The proposed health decision-making method takes into account each user's health condition so that the method has more precision than an existing recommendation method. In addition, the proposed method brings about better evaluation results than a general user-by-user health context information based recommendation method. CONCLUSION By recommending the food products related to users' chronic diseases through the proposed hybrid clustering, it is possible to help out their healthcare. In addition, by letting users receive satisfying feedback flexibly, it is possible to improve their dietary habits.

Due to our rapidly aging society, there has been an increasing need for preventive management of chronic diseases and management of individual health conditions. Among the chronic diseases, diabetes has become one of the most important illnesses in the modern era, and there is an increasing number of diabetes patients in all age groups. With increasing concern for ubiquitous health care services and the developing information technology, there has been a rapid increase in the convenience of preventive management for various types of disease and health conditions. In addition, demand for the management of chronic diseases by using internet of things devices has been increasing. In this study, a personal health record (PHR)-based diabetes index service model is suggested by analyzing lifestyles. For the method suggested in this study, a PHR-based mobile service for users is provided while developing the diabetes index model using lifestyle analysis. By utilizing the Korea National Health and Nutrition Examination Survey provided by the Ministry of Health and Welfare as a fundamental resource, this study analyzes the correlation and differentiation of a non-diabetic group and a diabetic group. For the selection of items in correlation, a multivariate analysis algorithm is used. A PHR related to health behavior and dietary habits in the same pattern is collected from users according to pre-chosen items. The Minkowski distance formula is used for the algorithm for PHR predictive analysis for users and chosen items. Resources on health conditions and dietary habits in the same pattern are collected from users, while calculating similarities among diabetes patients. By using the weight of calculated similarity, a diabetes index is derived. In addition, a management information service for preventing diabetes is offered through derived resources. A mobile application is developed from a smart health care platform to provide services at any time and any place. The mobile application's interface is configured for users to conveniently enter the PHR and check their health conditions on a real-time basis. It is feasible for each user to check the results from feedback by using the PHR. This quantifies the diabetes index and provides information about desirable health behavior and dietary habits related to diabetes at the same time. Users are able to conveniently apply them in real life with the provided services. Therefore, it is available to improve the health conditions of users and to prevent disease.

In processing voice with environment noise, the noise must be eliminated to improve the vocabulary recognition rate. In this process, noise elimination and feature extraction for model-estimate technologies are utilized. Concerning these noise-elimination and model-estimate technologies, the most important part is to estimate mixed noise in the source signal and eliminate it. In a vocabulary recognition system, if unexpected noise appears in the signal, or if quantization noise is basically added to digital signals, the source signal is changed or damaged, which decreases the recognition rate. If a source signal is transformed or changed by being mixed with diverse kinds of noise, the hidden Markov model (HMM) is used for effective noise elimination. The HMM forms a model by extracting features to flexibly respond to diverse vocabulary changes found in voice and text, etc. The method is applicable to data changing over time, and can establish a more effective model as the number of parameters constituting the model grows larger. It can provide a robust model estimate by using a parameter set for structured models. HMM-based vocabulary recognition shows discriminating distribution of recognition probability regarding recognition vocabulary models, and has lower computational complexity for recognition. But it produces a relatively lower recognition rate. To solve that problem, a vocabulary recognition-model optimization method is proposed based on a similar phoneme---recognition process and efficient feature extraction. In vocabulary recognition, a similar phoneme---recognition process is applied to HMM to recognize models adjacent to the model group. Efficient feature extraction is used to optimize the recognition model to enhance the recognition rate. For vocabulary composition, a Gaussian-mixture feature-extraction model is optimized and used as a vocabulary recognition model. Then, it is processed with similar-phoneme recognition regarding the vocabulary recognition model.

As augmented reality---related technologies become commercialized due to requests for 3D content, they are developing a pattern whereby users utilize and consume the realism and reality of 3D content. Rather than using absolute position information, the pattern characteristics of gestures are extracted by considering body-proportion characteristics around the shoulders. Even if performing the same gesture, position coordinate values of the skeleton measured by a motion recognition sensor can vary, depending on the length and direction of the arm. In this paper, we propose a 3D human-gesture interface for fighting games using a motion recognition sensor. Recognizing gestures in the motion recognition sensor environment, we applied the gestures to a fighting action game. The motion characteristics of gestures are extracted by using joint information obtained from the motion recognition sensor, and 3D human motion is modeled mathematically. Motion is effectively modeled and analyzed with a method of expressing it in space via principal component analysis and then matching it with the 3D human-gesture interface for new input. Also, we propose an advanced pattern matching algorithm as a way to reduce motion constraints in a motion recognition system. Finally, based on the results of motion recognition, an example used as the interface of a 3D fight action game is presented. By obtaining high-quality 3D motion, the developed technology provides more realistic 3D content through real-time processing technology.

As IT convergence technique develops, medical technology and apparatus are being modernized opening the era that we can obtain variable information easily anywhere, anytime thanks to wireless communication developed, further. These social changes enabled us to obtain information related to health more efficiently. Modern society is rapidly aging and more people experience chronic diseases because of their wrong eating habit, obesity and insufficient exercise. Thus a demand for health improvement and management at a certain term is increasing rather than complete therapy. Previously, major medical institutions managed personal medical history regarding patients mainly in health management but it is not changing its method to self-utilization and management by individual patient as of now along with medical institutions as fusion technology develops, and individual health record information can easily be checked anywhere, anytime through personal health record (PHR) platform. Unlike developing speed of related technology, however, there is a limitation in expansion, development of individual health record service, personal information security currently. In this paper, we propose mobile service regarding life style improvement targeting high risk chronic diseases based on PHR platform. PHR platform determines high blood pressure, diabetes, hyperlipidemia diseases which are three main chronic diseases using users' data and can monitor chronic diseases in portable mobile device. Also, the service provides by organically, mutually connected form through feedback towards input from health states of users in mobile device. By proposing contents about service based on efficient individual health record through mobile device that maximized transportability based on PHR platform, proposed method will contribute to industry development and activation of application service development of individual health record. Increase in consistency and reliability through standardization of afterwards health management service is expected to contribute to reduction in social cost and improvement of national health being the basis to realize communication activation of health record between medical institutions, efficient management and education of patients, reduction in dual examinations.

As an interest in health and disease has increased, medical service has changed to prevention of disease and health care from treatment oriented service. Medical service industry is creating various services and added value for promotion of health. Aging, extension of life expectancy, increase in lifestyle and income growth have brought about a change in paradigm of medical service which led smart health to become an important issue. Smart health caused medical service for promotion of health to change into remote medical treatment that uses personal health record from medical service which has been provided by mainly large hospitals. Medical service for promotion of health has developed into u-Healthcare which monitors condition of health in everyday life. This enabled problems of time and space constraints that occur in medical service for promotion of health that requires a medical doctor to examine bio-signal related information of a patient while facing a patient to be solved. It is difficult for a remote medical treatment to care for chronic patients who require a care of lifestyle because it focuses on treating specific diseases. As a remote medical treatment does not provide innovative medical service and it only delivers general bio information on a patient to a medical doctor remotely, remote medical open platform is needed. Thus, in this paper, we proposed a PHR open platform based smart health services using the distributed object group framework. A PHR open platform based smart health system is distributed object group framework based smart health service for managing chronic diseases. When Medical WBAN sensor uses multi-channel in transmitting data, emergency data is very important in patient's life, smart health environment is built using distributed network considering importance according to data. As WBAN sensor is very different from other networks in terms of application, architecture and density of development, it is important for WBAN sensor to be combined with external network. High quality of service of integrated network as well as link connectivity should be maintained. Since automatic diagnosis function should be reinforced in order for remote diagnosis service to be provided, integration of each small unit system and model design are important. Therefore, smart health network environment that makes the most of performance of distributed network based on automation technique and distributed agent for optimum design of system is built.

Noise-elimination technology is used to eliminate noise, including environmental noise, from voice signals in order to increase voice recognition rates. Noise estimation is the most important factor in noise-elimination technology. One of the effective estimation methods is voice activity detection, which is based on the statistical properties of noise and voice. This method is a way of estimating noise using the statistical properties of both noise and voice, which have an independent Gaussian distribution. In cases of severe differences in a statistical property, like white noise, the method is very reliable but limited to signals having a low signal-to-noise ratio (SNR) or having speech shape noise, which has statistical properties similar to voice signals. Methods to increase the voice recognition rate suffer from decreasing voice recognition performance due to distortion of the voice spectrum and to missing voice frames, because noise remains if there has been incorrect estimation of the noise. Degradation in voice recognition performance emerges in the differences between the model training environment and the voice recognition environment. In order to decrease environmental discordance, various silence feature normalization methods are used. Existing silence feature normalization suffers from degradation of recognition performance because the classification accuracy for the voiced and unvoiced signals decreases by an increasing energy level in the silence section of a low SNR. This paper proposes a robust voice characteristic detection method for noisy environments using feature extraction and unvoiced feature normalization for a classification relative to the voiced and unvoiced signals. The suggested method constitutes a model for recognition by extracting the characteristics for classification of the voiced and unvoiced signals in a high SNR environment. Also, the model affects noise for voice characteristics less, and recognition performance improves by using the Cepstrum feature distribution property of voiced and unvoiced signals with a low SNR. The model was checked for its ability to improve recognition performance relative to the existing method based on recognition experiment results.

This study proposes a continuous gesture recognition using information about features of motions obtained from continuous gesture images through higher order correlation feature coefficient and principal component analysis (PCA). The proposed method, first, separates two-dimensional silhouette gesture region from a continuous input image that includes a human body image. Information about 35 features are extracted using the higher order local auto correlation coefficient in the divided image and a low-dimensional gesture space is composed using PCA. The model feature value reflected in the gesture space is composed of symbols of certain conditions through clustering algorithm so as to be used as the input symbol of a hidden marker model, and a random input motion is recognized as the relevant gesture model with the highest probability value. The proposed method has less computation workload than the previous geometric feature-based method or appearance-based method and shows a high recognition rate with using the minimum information, so it is very suitable for the construction of a real-time system. In addition, the recognized gesture can be used input data of various applications like dynamic motion type game by mapping to an operating system. And the human motions can be used the objects of observation, such as smart home and behavior analysis in special spaces.