Search

To improve the accuracy of user implicit rating prediction, we combine the traditional latent factor model (LFM) and bidirectional gated recurrent unit neural network (BiGRU) model to propose a hybrid model that deeply mines the latent semantics in the unstructured content of the text and generates a more accurate rating matrix. First, we utilize the user’s historical behavior (favorites records) to build a user rating matrix and decompose the matrix to obtain the latent factor vectors of users and literature. We also apply the BERT model for word embedding of the research papers to obtain the sequence of word vectors. Then, we apply the BiGRU with the user attention mechanism to mine the research paper textual content and to generate the new literature latent feature vectors that are used to replace the original literature latent factor vectors decomposed from the rating matrix. Finally, a new rating matrix is generated to obtain users’ ratings of noninteractive research papers and to generate the recommendation list according to the user latent factor vector. We design experiments on the real datasets and verify that the research paper recommendation model is superior to traditional recommendation models in terms of precision, recall, F1-value, coverage, popularity and diversity.

This companion paper is to support the replication of paper "Campus3D: A Photogrammetry Point Cloud Benchmark for Outdoor Scene Hierarchical Understanding", which was presented at ACM Multimedia 2020. The supported paper's main purpose was to provide a photogrammetry point cloud-based dataset with hierarchical multilabels to facilitate the area of 3D deep learning. Based on this provided dataset and source code, in this work, we build a complete package to reimplement the proposed methods and experiments (i.e., the hierarchical learning framework and the benchmarks of the hierarchical semantic segmentation task). Specifically, this paper contains the technical details of the package, including file structure, dataset preparation, installation package, and the conduction of the experiment. We also present the replicated experiment results and indicate our contributions to the original implementation.

We introduce a comprehensive benchmark for local features and robust estimation algorithms, focusing on the downstream task -- the accuracy of the reconstructed camera pose -- as our primary metric. Our pipeline's modular structure allows easy integration, configuration, and combination of different methods and heuristics. This is demonstrated by embedding dozens of popular algorithms and evaluating them, from seminal works to the cutting edge of machine learning research. We show that with proper settings, classical solutions may still outperform the perceived state of the art. Besides establishing the actual state of the art, the conducted experiments reveal unexpected properties of Structure from Motion (SfM) pipelines that can help improve their performance, for both algorithmic and learned methods. Data and code are online https://github.com/vcg-uvic/image-matching-benchmark, providing an easy-to-use and flexible framework for the benchmarking of local features and robust estimation methods, both alongside and against top-performing methods. This work provides a basis for the Image Matching Challenge https://vision.uvic.ca/image-matching-challenge., Comment: Added: KeyNet-SOSNet, AffNet-HardNet, TFeat, MKD from kornia

Summary Objectives : Analyze papers published in 2019 within the medical natural language processing (NLP) domain in order to select the best works of the field. Methods : We performed an automatic and manual pre-selection of papers to be reviewed and finally selected the best NLP papers of the year. We also propose an analysis of the content of NLP publications in 2019. Results : Three best papers have been selected this year including the generation of synthetic record texts in Chinese, a method to identify contradictions in the literature, and the BioBERT word representation. Conclusions : The year 2019 was very rich and various NLP issues and topics were addressed by research teams. This shows the will and capacity of researchers to move towards robust and reproducible results. Researchers also prove to be creative in addressing original issues with relevant approaches.

In the paper we present an approach to the automatic detection and identification of important elements in paper documents. This includes stamps, logos, printed text blocks, signatures and tables. Presented approach consists of two stages. The first one includes object detection by means of AdaBoost cascade of weak classifiers and Haar-like features. Resulting image blocks are, at the second stage, subjected to verification based on selected features calculated from recently proposed low-level descriptors combined with certain classifiers representing current machine-learning approaches. The training phase, for both stages, uses bootstrapping, i.e., integrative process, aiming at increasing the accuracy. Experiments performed on large set of digitized paper documents showed that adopted strategy is useful and efficient.

This paper presents two approaches for monitoring the performance of the paper making process. The first uses a neural network vibration-based condition monitoring system for providing advance warning of press felt problems. The characteristics of this system include spectral analysis through acceleration enveloping, a peak detection algorithm as a feature extractor and a neural network classifier. The second approach makes use of multivariate statistical techniques. This involves the generation of a multivariate statistical representation of nominal process behaviour. Process malfunctions are then identified by deviations from the developed model. Both approaches are successfully applied to an industrial paper machine.

Exploring massive mobile data for location-based services becomes one of the key challenges in mobile data mining. In this paper, we investigate a problem of finding a correlation between the collective behavior of mobile users and the distribution of points of interest (POIs) in a city. Specifically, we use large-scale cell tower data dumps collected from cell towers and POIs extracted from a popular social network service, Weibo. Our objective is to make use of the data from these two different types of sources to build a model for predicting the POI densities of different regions in the covered area. An application domain that may benefit from our research is a business recommendation application, where a prediction result can be used as a recommendation for opening a new store/branch. The crux of our contribution is the method of representing the collective behavior of mobile users as a histogram of connection counts over a period of time in each region. This representation ultimately enables us to apply a supervised learning algorithm to our problem in order to train a POI prediction model using the POI data set as the ground truth. We studied 12 state-of-the-art classification and regression algorithms; experimental results demonstrate the feasibility and effectiveness of the proposed method.

We conducted a reliability study comparing single data entry (SE) into a Microsoft Excel spreadsheet to entry using the existing forms (EF) feature of the Teleforms software system, in which optical character recognition is used to capture data off of paper forms designed in non-Teleforms software programs. We compared the transcription of data from multiple paper forms from over 100 research participants representing almost 20,000 data entry fields. Error rates for SE were significantly lower than those for EF, so we chose SE for data entry in our study. Data transcription strategies from paper to electronic format should be chosen based on evidence from formal evaluations, and their design should be contemplated during the paper forms development stage.

plays a vital role in research area in the field of data mining. Clustering is a process of partitioning a set of data in a meaningful sub classes called clusters. It helps users to understand the natural grouping or cluster from the data set. It is unsupervised classification that means it has no predefined classes. This paper presents a study of various partitioning techniques of clustering algorithms and their relative study by reflecting their advantages individually. Applications of cluster analysis are Economic Science, Document classification, Pattern Recognition, Image Processing, text mining. No single algorithm is efficient enough to crack problems from different fields. Hence, in this study some algorithms are presented which can be used according to one's requirement. In this paper, various well known partitioning based methods - k-means, k-medoids and Clarans - are compared. The study given here explores the behaviour of these three methods.

In the paper “A hybridized K-means clustering approach for high dimensional dataset” Dash, Mishra, Rash and Acharya have presented a new version of the k-means algorithm. In it, principal components analysis (PCA) was used before applying the kmeans algorithm with a new initialization method. The authors compare the results obtained by using the HKMCA and PCA with the results of the original k-means, but a direct comparison is not valid as this paper shows. Keywords: Cluster analysis, K-means Algorithm, Principal Component Analysis, Hybridized K-means algorithmInternational Journal of Engineering, Science and Technology Vol. 6, No. 1, 2014, pp. 20-26

Steganography is the skill of hiding the existence of data in other transmission medium to attain secret communication. It does not restore cryptography but quite boost the security using its abstruse features. In this paper we have surveyed on a Steganography and cryptography techniques which provide highly secure skin tone data hiding. Biometric characteristic used to apply steganography is skin tone region of images. Here important data is implanted within skin region of image which will give an outstanding secure location for data hiding. For this skin tone detection is need to be performed. Different steps of data hiding can be applied by cropping an image interactively. Cropping of an image improved security than hiding data without cropping the whole image, so cropped region works as a key at decoding region. Cryptography algorithm is used to convert the secret messages to an unreadable form before embedding; which provides a strong backbone for data security. This survey paper focuses on illuminating the technique to secure data or message with authenticity and non repudiation. So with this object oriented steganogaphy we track skin tone objects in image with the higher security and satisfactory PSNR .Modern steganography’s goal is to keep its mere presence undetectable.

amount information of moving objects on road network is being collected with the help of various recent technologies. The tracking of these moving objects on road networks is becoming important because of it's application in various areas. Classification has been used for classifying various kinds of data sets like graph, text documents. However, there is a lack of study on data like trajectories on road networks. Data mining techniques, especially, sequential pattern mining can be used to extract frequent spatio-temporal patterns. Again it needs to confine the length of sequential patterns to ensure high efficiency. After extracting frequent sequential patterns in trajectories, classification can be applied to classify patterns which provide useful information in applications such as city and transportation planning, road construction, design, and maintenance, marketing sector. In this paper, whole pattern matching query concept is adopted after the classification to find total traffic volume on given trajectory edge. At the same time, user can find number of vehicles moving in one as well as in both directions on that particular trajectory. Keywordssequential patterns, frequent pattern based classification, location based services, pattern matching.

Data stream is massive, fast changing and infinite in nature. It is very natural that large amount of unlabeled data and small amount labeled are available in data stream environments. Storing and labeling all data is considered expensive and impractical. The objective is to label small portion of stream data and analyze data online without storing it. Concept drift, concept evolving, stream evolving is also the major challenging problem occurs while working with data stream. Online data stream active learning is needed to tackle these problems. Classification and clustering are two technical areas that are widely used to extract pattern from the large data stream, from that a classification model must endlessly adapt itself to the most recent concept. Hence, this paper gives the overview of various ensemble based classification algorithm techniques in the field of data stream mining.

This article explores the application of image processing techniques in recyclable waste paper sorting. In recycling, waste papers are segregated into various grades as they are subjected to different recycling processes. Highly sorted paper streams facilitate high quality end products and save processing chemicals and energy. From 1932 to 2009, different mechanical and optical paper sorting methods have been developed to fill the paper sorting demand. Still, in many countries including Malaysia, waste papers are sorted into different grades using a manual sorting system. Because of inadequate throughput and some major drawbacks of mechanical paper sorting systems, the popularity of optical paper sorting systems has increased. Automated paper sorting systems offer significant advantages over human inspection in terms of worker fatigue, throughput, speed, and accuracy. This research attempts to develop a smart vision sensing system that is able to separate the different grades of paper using first-order features. To construct a template database, a statistical approach with intra-class and inter-class variation techniques are applied to the feature selection process. Finally, the K-nearest neighbor (KNN) algorithm is applied for paper object grade identification. The remarkable achievement obtained with the method is the accurate identification and dynamic sorting of all grades of papers using simple image processing techniques.

Problem Statement: Almost all of the modern biomedical equipments that record the biopotential actions have digital output but the paper chart records are a must. The volume of these records is significant and increasing rapidly. Keeping a bio-signal chart of a patient make it easy for quick assessment but still create problems in the essence of data storage, archiving , data interchange and communications. Approach: As a solution to all these problems is to convert these paper records to digital form. In this study, a method for bio-potential signal extraction from single or multi-channel paper recorded charts using image processing techniques is developed. Results: After scanning the paper charts and converting them into images using a commercial scanner, the developed algorithm is applied to eliminate the background of the scanned paper chart from any recording device single channel or multi-channel using binary neighborhood morphological operations then converting the extracted waveform image into quantized values representing the waveform recorded paper chart. The extracted signal is then filtered to remove the high frequency effects that result from the morphological operations. A correlation and frequency analysis procedure is then conducted to verify the result against known sampled waveform. Conclusion: A chart paper conversion to digital values that represent only the values for the biopotential waveform and eliminating the other irrelevant information has been achieved. These resulted in a less space occupation of patient records and make it easy for data further processing and manipulations.

Data hiding technique can facilitate security and the safe transmission of important information in the digital domain, which generally requires a high embedding payload and good stego image quality. Recently, a steganographic framework known as wet paper coding has been utilized as an effective strategy in image hiding to achieve the requirements of high embedding payload, good quality and robust security. In this paper, besides employing this mechanism as a fundamental stage, we take advantage of two novel techniques, namely, an efficient n-indicator and a fuzzy edge detector. The first is to increase the robustness of the proposed system to guard against being detected or traced by the statistics methods while allowing the receiver without knowledge of secret data positions to retrieve the embedded information. The second is to improve the payload and enhance the quality of stego image. The experimental results show that our proposed scheme outperforms its ability to reduce the conflict among three steganography requirements.

This paper describes the possibilities of variable selection in large-scale industrial systems. It introduces knowledge-based, data-based and model-based methods for this purpose. As an example, Case-Based Reasoning application for the evaluation of the web break sensitivity in a paper machine is introduced. The application uses Linguistic Equations approach and basic Fuzzy Logic. The indicator combines the information of on-line measurements with expert knowledge and provides a continuous indication of the break sensitivity. The web break sensitivity defines the current operating situation at the paper mill and gives new information to the operators. Together with information of the most important variables this prediction gives operators enough time to react to the changing operating situation.

This is the post-print version of the final paper published in Advanced Engineering Informatics. The published article is available from the link below. Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. Copyright @ 2013 Elsevier B.V. The purpose of this research is twofold: first, to undertake a thorough appraisal of existing Input Variable Selection (IVS) methods within the context of time-critical and computation resource-limited dimensionality reduction problems; second, to demonstrate improvements to, and the application of, a recently proposed time-critical sensitivity analysis method called EventTracker to an environment science industrial use-case, i.e., sub-surface drilling. Producing time-critical accurate knowledge about the state of a system (effect) under computational and data acquisition (cause) constraints is a major challenge, especially if the knowledge required is critical to the system operation where the safety of operators or integrity of costly equipment is at stake. Understanding and interpreting, a chain of interrelated events, predicted or unpredicted, that may or may not result in a specific state of the system, is the core challenge of this research. The main objective is then to identify which set of input data signals has a significant impact on the set of system state information (i.e. output). Through a cause-effect analysis technique, the proposed technique supports the filtering of unsolicited data that can otherwise clog up the communication and computational capabilities of a standard supervisory control and data acquisition system. The paper analyzes the performance of input variable selection techniques from a series of perspectives. It then expands the categorization and assessment of sensitivity analysis methods in a structured framework that takes into account the relationship between inputs and outputs, the nature of their time series, and the computational effort required. The outcome of this analysis is that established methods have a limited suitability for use by time-critical variable selection applications. By way of a geological drilling monitoring scenario, the suitability of the proposed EventTracker Sensitivity Analysis method for use in high volume and time critical input variable selection problems is demonstrated. EU

In order to keep up with the requirements of College English Diagnostic Test System developed by our team, after analyzing the pros and cons of current test paper generating algorithms, the MCKP algorithm, which is based on maximum coverage of knowledge points (MCKP), is proposed. In this paper, the knowledge hierarchy of College English Test (CET4) of China is constructed to provide the foundation for MCKP. Then given three main parameters, including the required items number, the mastery rates of knowledge-point and the testing frequencies of knowledge points, the theory and implementation of the MCKP algorithm is elaborated. At last, experiments indicate relatively high success rate and effectiveness of test paper generating process. Therefore, MCKP algorithm can provide an essential improvement in test paper generating process. Keywordsmaximum covarage of knowledge points (MCKP); knowledge hierarchy of CET-4;mastery rates ;testing frequencies; success rate

Feature selection determines relevant features in the data. It is often applied in pattern classification, data mining, as well as machine learning. A special concern for feature selection nowadays is that the size of a database is normally very large, both vertically and horizontally. In addition, feature sets may grow as the data collection process continues. Effective solutions are needed to accommodate the practical demands. This paper concentrates on three issues: large number of features, large data size, and expanding feature set. For the first issue, we suggest a probabilistic algorithm to select features. For the second issue, we present a scalable probabilistic algorithm that expedites feature selection further and can scale up without sacrificing the quality of selected features. For the third issue, we propose an incremental algorithm that adapts to the newly extended feature set and captures `concept drifts' by removing features from previously selected and newly added ones. We expect that research on scalable feature selection will be extended to distributed and parallel computing and have impact on applications of data mining and machine learning.

In this paper a new technique is proposed to improve the recognition ability and the transaction speed to classify the Japanese and US paper currency. Two types of data sets, time series data and Fourier power spectra, are used in this study. In both cases, they are directly used as inputs to the neural network. Furthermore, we also refer a new evaluation method of recognition ability. Meanwhile, a technique is proposed to reduce the input scale of the neural network without preventing the growth of recognition. This technique uses only a subset of the original data set which is obtained using random masks. The recognition ability of using large data set and a reduced data set are discussed. In addition to that the results of using a reduced data set of the Fourier power spectra and the time series data are compared. >

In the last few years, many image processing techniques have been presented in order to perform different brain tumor detection tasks. These cover Content-Based Retrieval Technique, Component Labelling Algorithm, Fuzzy C-Mean Algorithm. There is fast growth of image processing available in last few years; image segmentation have work together to get extraction of meaningful data into useful information. Image processing is used to finding applicable and efficient information from data available. With this survey paper we focus on literature survey of segmentation based image processing model as contain efficient performance.

To study the performance of rock cutting trenchers, data on the excavation and tool consumption rate of one type of trencher, the Vermeer T-850, were gathered on 16 sites. The data assembled were compared with the rock characteristics by studying the trench geology and performing rock engineering tests on samples. This study aims at more reliable predictions, by developing better methods to handle the data, which are commonly of an imprecise nature. To reach this goal, fuzzy set theory has been selected and successfully implemented. A Fuzzy Expert System model has been developed to predict the bit consumption and the excavation rate of the T-850 trencher. The results obtained so far are promising and the model is in the verification stage.

Background In recent years high throughput methods have led to a massive expansion in the free text literature on molecular biology. Automated text mining has developed as an application technology for formalizing this wealth of published results into structured database entries. However, database curation as a task is still largely done by hand, and although there have been many studies on automated approaches, problems remain in how to classify documents into top-level categories based on the type of organism being investigated. Here we present a comparative analysis of state of the art supervised models that are used to classify both abstracts and full text articles for three model organisms. Results Ablation experiments were conducted on a large gold standard corpus of 10,000 abstracts and full papers containing data on three model organisms (fly, mouse and yeast). Among the eight learner models tested, the best model achieved an F-score of 97.1% for fly, 88.6% for mouse and 85.5% for yeast using a variety of features that included gene name, organism frequency, MeSH headings and term-species associations. We noted that term-species associations were particularly effective in improving classification performance. The benefit of using full text articles over abstracts was consistently observed across all three organisms. Conclusions By comparing various learner algorithms and features we presented an optimized system that automatically detects the major focus organism in full text articles for fly, mouse and yeast. We believe the method will be extensible to other organism types.

This chapter studies and analyzes Memetic Differential Evolution (MDE) Frameworks for designing digital filters, which aim at detecting paper defects produced during an industrial process. MDE Frameworks employ the Differential Evolution (DE) as an evolutionary framework and a list of local searchers adaptively coordinated by a control scheme. Here, three different variants of MDE are taken into account and their features and performance are compared. The binomial explorative features of the DE framework in contraposition to the exploitative features of the local searcher are analyzed in detail in light of the stagnation prevention problem, typical for the DE. Much emphasis in this chapter is given to the various adaptation systems and to their applicability this image processing problem.

The Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD) has been held every year since 1997. PAKDD 2008, the 12th in the series, was held in Osaka, Japan on May 20–23, 2008. PAKDD is a leading international conference in the area of data mining. It provides an international forum for researchers and industry practitioners to share their new ideas, original research results and practical development experiences from all KDD-related areas including data mining, data warehousing, machine learning, databases, statistics, knowledge acquisition, automatic scientific discovery, data visualization, causal induction and knowledge-based systems. We received a total of 312 research papers from34 countries and regions inAsia, Australia, North America, South America, Europe and Africa. Only approximately 11.9% of the these submissions were accepted as long papers, 12.8% of them were accepted as regular papers, and 11.5% of them were accepted as short papers upon rigorous reviews by two or three reviewers, discussions by the reviewers under the supervision of an area chair and judgment by the Program Committee Co-chairs. The highest rated papers in the review process were further evaluated by the Program Committee Co-chairs, and two papers were selected for the Best Paper Award and the Best Paper Runner-up Award. In addition to these award-winning papers, two outstanding papers were also nominated for this special issue. The authors of the four selected papers were asked to substantially extend the contents of their papers after the

Objective: We present a Matlab-based tool to convert electrocardiography (ECG) information from paper charts into digital ECG signals. The tool can be used for long-term retrospective studies of cardiac patients to study the evolving features with prognostic value. Methods and procedures: To perform the conversion, we: 1) detect the graphical grid on ECG charts using grayscale thresholding; 2) digitize the ECG signal based on its contour using a column-wise pixel scan; and 3) use template-based optical character recognition to extract patient demographic information from the paper ECG in order to interface the data with the patients' medical record. To validate the digitization technique: 1) correlation between the digital signals and signals digitized from paper ECG are performed and 2) clinically significant ECG parameters are measured and compared from both the paper-based ECG signals and the digitized ECG. Results: The validation demonstrates a correlation value of 0.85–0.9 between the digital ECG signal and the signal digitized from the paper ECG. There is a high correlation in the clinical parameters between the ECG information from the paper charts and digitized signal, with intra-observer and inter-observer correlations of 0.8–0.9 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$({\rm p}