Search

© 2013 IEEE. With the rapid growth of scientific publications, it is hard for researchers to acquire appropriate papers that meet their expectations. Recommendation system for scientific articles is an essential technology to overcome this problem. In this paper, we propose a novel low-rank and sparse matrix factorization-based paper recommendation (LSMFPRec) method for authors. The proposed method seamlessly combines low-rank and sparse matrix factorization method with fine-grained paper and author affinity matrixes that are extracted from heterogeneous scientific network. Thus, it can effectively alleviate the sparsity and cold start problems that exist in traditional matrix factorization based collaborative filtering methods. Moreover, LSMFPRec can significantly reduce the error propagated from intermediate outputs. In addition, the proposed method essentially captures the low-rank and sparse characteristics that exist in scientific rating activities; therefore, it can generate more reasonable predicted ratings for influential and uninfluential papers. The effectiveness of the proposed LSMFPRec is demonstrated by the recommendation evaluation conducted on the AAN and CiteULike data sets.

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.

As the connectivity of autonomous vehicles keeps growing, it is an accepted fact that they are even more vulnerable to malicious cyber-attacks. Recently, sensor spoofing has become an emerging attack that can compromise vehicle safety as vehicles are equipped with more sensors. Thus, it is critical to validate the sensor readings before utilizing them for future actions. In this paper, we develop HDAD, a hyperdimensional computing-based anomaly detection method. Hyperdimensional computing (HDC) is an emerging brain-inspired computing paradigm that mimics the brain cognition and leverages hyperdimensional vectors with fully distributed holographic representation and (pseudo)randomness. The key idea of HDAD is to use HDC to build encoder and decoder to reconstruct the sensor readings. The anomalous data typically have comparatively higher reconstruction errors than normal sensor readings. We explore three different metrics to measure the reconstruction error including mean squared error, mean absolute error, and cosine similarity. Using a real-world vehicle sensor reading dataset, we demonstrate the feasibility and efficacy of HDAD, opening the door for a new set of anomaly detection algorithm design.

Facing the difficulty and inefficiency of creating and maintaining High-Definition (HD) maps in our commercial deployments, we have developed an edge-based crowdsourcing task distribution framework for HD Map in autonomous driving. Our key observation is that: HD map data crowdsourcing exhibits the diminishing marginal utility thus there exists an inflection point for maximum utility, meanwhile its premature convergence of utility will leave some map updates not notified in time. Based on this observation, we develop a periodic crowdsourcing task distribution framework. It discretizes the demands for collecting source data into different periods and uses an optimal stopping rule to terminate the data collection for the maximum crowdsourcing utility. The experimental results verify that our crowdsourcing framework can achieve high time coverage and high efficiency with lower cost.

With the huge amount of published research papers, retrieving relevant information is a difficult task for any researcher. Effective clustering algorithms can help improve and simplify the retrieval process. Here, we propose an approach for automatic clustering for text document using a Self-Organizing Map (SOM). It is one of unsupervised artificial neural network that widely used for data analysis, data compression, clustering, and data mining. The quality and accuracy of a SOM algorithm depends on the selection of values for some of its parameters which are its initial learning rate, SOM matrix dimensions, and the number of iterations. Best values are typically selected using trial and error; however, in the current paper we suggest a more systematic approach to parameters optimization using the genetic algorithm. The proposed method is applied to cluster 3 scientific papers datasets using their keywords. Similar research papers were mapped closer to each other. Clustering results were validated using the Dunn index.

In the present work, the 48 input-output data points of the paper mill process have been considered with the desired machine speed as the input and required total head as the output. The data has been taken from the literature. Two clustering techniques are applied on this data using MATLAB. One is the FCM (fuzzy C-means clustering), another is the subtractive clustering. In present work, initial FIS (fuzzy inference system) developed by subtractive clustering is further optimized to improve its performance, and finally their responses are compared.

BBoxDB is a distributed key-bounding-box-value store which is capable of handling large scale $n-$dimensional data. Unlike existing key-value stores, each value is stored together with a bounding box which describes the location of the value in an $n-$dimensional space. BBoxDB splits large datasets automatically and spreads them across a cluster of nodes. The software works primarily on the bounding boxes of the stored data; operations like range queries or joins take only the bounding boxes of the data into consideration. In version 0.9.1, we implemented support for user defined filters (UDF) in BBoxDB; UDFs operate on the real values of the data. In this paper, we describe this novel feature for the first time. As an example, a UDF is developed in this paper which is capable of handling GeoJSON encoded data. The UDF consists only of a few lines of code and turns the generic distributed datastore BBoxDB into a specialized system that is capable of processing large scale GeoJSON encoded datasets. Operations like spatial joins or range queries on the real geometries of the data become possible. During our demonstration, we show the implementation of the UDF and perform range queries and spatial joins. The used dataset contains spatial data of the whole world and is obtained from the Open Street Map project. For visualization, the GUI of BBoxDB was extended in a way that spatial queries can easily be executed. Using the GUI, query results can be interactively explored; they are shown as an overlay on a map dynamically fetched from the Open Street Map project.

Privacy can be revealed from many sources, since data is available in papers and digital forms. Many techniques available to infer some private data based on these available sources, for example, data mining techniques. In this paper, it is investigated whether it is possible to derive some private information from shopping receipts, which are paper traces of people activities. Since people do not think that shopping receipts may harm them, usually they do not keep these shopping receipts safely. The techniques used in this research are data mining techniques; classification and clustering algorithm. In this first stage of research, the aim of this research is to determine whether it is possible to classify shopping data to correct shopper profiles, and to separate a set of shopping data into two meaningful groups, according to the shopping profiles. From the results, it can be concluded that it is possible to predict to which profile the data belongs to, with the percentage of data correctly classified higher than 97%. It also can be concluded that it is possible to separate the shopping receipts data into two groups, where each group represents each profile.

Latent Semantic Analysis uses Singular Value Decomposition (SVD) to effectively retrieve relevant information from the information corpus. However, LSA has a high computational cost. In order to address this aspect, it is proposed to filter only those words carrying high semantic importance. The aim is to improve the execution time of semantic space construction and dimensionality reduction. We present how the use of data filter can effectively meet the proposed goals in comparison to baseline method of performing recommendations. The proposed system was assessed over a dataset of 80 articles (Titles and Abstracts). The results of the experiments show that the proposed system performed better in terms of elapsed time with an average precision of 85.54% (78.64% for baseline method) and an average recall of 92.96% (89.70% for baseline method).

Association Rule Hiding is the Data Mining method which is used for extracting hidden information from huge dataset. In our paper, Two approaches are introduced. In first approach FP Growth Algorithm is being presented that generate association rules efficiently and it reduces time of forming frequent item sets every time. In second approach we have tried to hide sensitive association rules by Genetic Algorithm. Generally from large databases frequent items are extracted by applying different algorithms. In this paper, we compare all the algorithms for extracting frequent items which are Association Rule Mining, Apriori Algorithm and FP growth Algorithm. We also compare Fuzzy Logic Algorithms and Genetic Algorithm for hiding sensitive association rules.

In this paper., the application of genetic algorithm in auto-generating test paper is studied. The linear discriminant analysis(LDA)is used to personally train the weight matrix of the objective function of the test paper. The optimal retention strategy and niche technology based on exclusion mechanism are combined to solve the problem that genetic algorithm is easy to “premature” convergence, and the parallel selection method is introduced to improve the running speed of the algorithm. In addition, the adaptive technique is used to set the appropriate exchange and mutation probability for genetic algorithm, and mountain climbing algorithm(MCA)is introduced to simulate the mutation process. Based on this, a new test paper algorithm(LNPAC-GA)is proposed., and is applied to the automatic test volume. Finally, an experiment is carried out by using the test question bank, in which the new algorithm is compared with the existing improved genetic algorithm (i.e. real coding, niche technology based on exclusion mechanism, adaptive technology to improve the genetic algorithm)which is accepted in the field of automatically generating test paper. Experiments show that the new algorithm can improve the speed and success rate of automatically generating test paper.

In this modern world, the correct identification of paper currency is an essential need for the automation of current systems. The automatic currency identification method is highly required to reduce the human power in recognizing the currency and its associated value. Sometimes, the damaged or blurred currency and the complex design of the currency make the task of identification much difficult. In this paper, a method is proposed that can identify paper currency by considering the identification marks given on the currency. Specifically, this study focuses on the identification of Pakistani paper currency. The proposed method makes use of the fundamental image processing techniques for differentiating the major denominations. Particularly, the proposed algorithm uses Otsu’s thresholding method and circulatory matrix criteria for locating the special identity marks given on the currency. The effectiveness of the proposed approach has been shown by performing various experiments on all of the major denominations of Pakistani paper currency with varying intensities. The Proposed method shows quite satisfactory results in terms of accuracy and efficiency.

Uncertainty quantification techniques have been widely used to model devices, circuits, and systems with fabrication process variations. However, most existing techniques assume that all random parameters are mutually independent or Gaussian correlated, which is rarely true in practice. How to handle non-Gaussian correlated random parameters is a fundamental and long-standing challenge. In this paper, we review existing techniques and point out their limitations. Then, we summarize our recent work to address the theoretical and computational challenges caused by non-Gaussian correlations.

Traditional signature-based malware detection techniques have been used for many years owing to their high detection rates and low false positive rates. However, signature-based techniques are ineffective as they do not detect new, polymorphic or transient malware. To beat weaknesses in signature-detection techniques, researchers have turned to behaviour-based ones, which create a malware behaviour profile by capturing malicious API calls throughout execution. In this context, API matching techniques use API calls to calculate similarities between malware. However, API concatenation techniques need significant API call process resources, which makes these methods slow owing to process quality, and thus cannot scale to large API call sequences outside the lab. In this paper, we review present malware detection strategies supported by API call sequences.

With the continual growth of low-cost and high-throughput DNA sequence technology, the scale and amount of next-generation sequencing (NGS) datasets are continually increasing in many genomics research areas. Shotgun metagenomics sequencing provides comprehensive information on microorganisms, based on complex samples of the ecosystem. Due to challenges of its scale and computational complexity, efficient sequence processing and analyzing tools are needed. In this paper, we propose a novel high-performance shotgun metagenomics sequence analysis engine for the task of sequence comparison. It includes two major components. First, a customized shifting database, which is optimized from any existing DNA sequence dataset. Second, a high-performance sequence computation algorithm that utilizes the customized shifting reference database and accelerates GPU parallel computing. We elaborated upon the efficiency and computational complexity of our proposed approach in an HPC server, which has eight Nvidia Tesla P100 GPUs. We also conducted a case study to detect viral sequences from patients' blood samples. Our experimental result shows that we obtain similar accuracy to the conventional BLAST method, but with a computational speed that is about twenty times faster.

Data storage in the Cloud became a very popular service. However, delegation of data management results in loss of control from user perspective, in particular regarding the real location where data are stored. Thus, data location verification in the Cloud is a challenging issue. Among the huge methods proposed to consider data location verification, this paper focuses on machine learning based methods, which use network Round Trip Times as main metric. In particular, it provides experimental results based on country-wide dataset collected through Grid'5000 platform. Results show the capacities of regression-based methods to support data location verification at specific accuracy depending on user requirements.

Defect prediction studies have proposed several data-driven approaches, and recently, this field has put more emphasis on whether the people factor is associated software defects. Developer metrics can capture experience, code ownership, coding skills and techniques, and commit activities. These metrics have so far been measured at a specified snapshot of the codebase although developer's knowledge on a source module could change over time. In this paper, we propose to measure periodic developer experience with regard to contextual knowledge on files and directories. We extract periodic experience metrics capturing the previous activities of developers on source files and investigate the explanatory effect of these metrics on defects. We also use activity-based (churn) metrics to observe the performance of both metric types on defect prediction. We used two large-scale open source projects, Lucene and Jackrabbit, for model evaluation. We calculate periodic developer experience metrics and churn metrics at two granularity levels: file level and commit level. We build the models using five popular machine learning algorithms in defect prediction literature. The models with the two best performing algorithms are assessed in terms of Precision, Recall, False Positive Rate, and F-measure. The set of metrics that explains software defects the best is also identified using correlation-based feature selection method. Results show that periodic developer experience metrics extracted at file level are good merits for defect prediction, accompanied with churn. When there is not enough data to extract the contextual knowledge of developers on source files, churn metrics play an important role on defect prediction.

Genetic network inference methods using random forests have shown promise. Some of the random-forest-based inference methods have an ability to analyze both time-series and static gene expression data. We think however that, as the gene expression levels at two adjacent measurements of a time-series data are often similar to each other, the gene expression levels at each measurement in the time-series data are less informative than those in the static data. On the basis of this idea, we proposed a new inference method that relies more on static gene expression data than time-series ones. Through the numerical experiments, we showed that the quality of the inferred genetic network is slightly improved by giving greater importance to static data than time-series ones. Although we develop the new method by modifying the random-forest-based inference method proposed by the authors, we could introduce the idea in this study into any inference method that is capable of analyzing both time-series and static gene expression data.

The online system is rapidly developing and widely used for companies to market their products. The products can always be more diverse and abundant. However, it creates difficulties for the company to provide recommendations to users about products which are suitable with users' interest. This condition encourages a recommendation system. One of the popular methods of this recommendation system is Collaborative Filtering (CF) by using rating-based and ranking-based approaches. Some of the ranking-based methods are Copeland method and Borda method. Both methods use a user-rating approach limited to the user preference profiling processes. Therefore, this research proposes the use of the user-rating to get the normalized rating frequency (NRF). Normalization process was done through the calculation of the frequency of a user-rating, which eventually generated a product ranking for recommendations to users. Experimental results of the NRF method can improve the performance of the recommendation system. This can be seen from the recommendations produced by the NRF method was more relevant in accordance with the wishes of the user, which is indicated by the average value of Normalized Discounted Cumulative Gain (NDCG) higher than Copeland and Borda methods. In addition, the NRF method has a faster computation time with a simpler algorithm than the Copeland and Borda methods.

Local differential privacy (LPD) is a distributed variant of differential privacy (DP) in which the obfuscation of the sensitive information is done at the level of the individual records, and in general it is used to sanitize data that are collected for statistical purposes. LPD has the advantage it does not need to assume a trusted third party. On the other hand LDP in general requires more noise than DP to achieve the same level of protection, with negative consequences on the utility. In practice, utility becomes acceptable only on very large collections of data, and this is the reason why LDP is especially successful among big companies such as Apple and Google, which can count on a huge number of users. In this talk, we propose a variant of LDP suitable for metric spaces, such as location data or energy consumption data, and we show that it provides a much higher utility for the same level of privacy. Furthermore, we discuss algorithms to extract the best possible statistical information from the data obfuscated with this metric variant of LDP.

In recent years, data mining had become very popular to provide a facility in various fields. Large amount of data elements are stored in the database. Various data mining algorithms are used to classify the data elements and perform some operations to find the best solution on the basis of some relational parameters. One of the automatic architecture designs in this research is to find the solution for examination in terms of level wise question paper. The entire examination process is a vital component for direct assessment of an individual learning. So, preparing a complete test paper and the setup is relatively necessary. Currently, the customary technique of making question paper has been handbook. In current scenario the question paper generation is a manual approach leading to unproductive at times owing to bias, repetition and security concerns. The current paper presents an automatic procedure of question paper group which can be modified, streamlined, synchronized and secured. Each task done by this system is automatic, such that storing space, bias and security is not an apprehension any longer. Earlier, the question paper was generated by concerned subject teacher manually and was very time consuming, man power was required and sometimes the question paper lacked accuracy. Outcome Based Education (OBE) designates what students will know and be intelligent to do, as they advance in a program. The information related to student learning was collected for question paper generation and assessment using OBE. This information can be used to predict students ability, advancements in education system, betterment in teaching method, future interest of student etc. Various data mining classifiers like oneR, ZeroR, J48, Naive Bayes, IBk are used for prediction of the course outcome. The comparative study contains some parameters like time, detection accuracy, classification error etc. to find the performance of the proposed architecture as compared to other techniques to verify the enhancements.

The enormous growth of information on the Internet makes finding information challenging and time consuming. Recommender systems provide a solution to this problem by automatically capturing user interests and recommending related information the user may also find interesting. In this paper, we present a novel recommender system for the research paper domain using a Dynamic Normalized Tree of Concepts (DNTC) model. Our system improves existing vector and tree of concepts models to be adaptable with a complex ontology and a large number of papers. The proposed system uses the 2012 version of the ACM Computing Classification System (CCS) ontology. This ontology has a much deeper structure than previous versions, which makes it challenging for previous ontology-based approaches to recommender systems. We performed offline evaluations using papers provided by ACM digital library for classifier training, and papers provided by CiteSeerX digital library for measuring the performance of the proposed DNTC model. Our evaluation results show that the novel DNTC model significantly outperforms the other two models: non-normalized tree of concepts and the vector of concepts models. Further, our DNTC model provides high average precision and reliable results when used in a context which the user has multiple interests and reads a large quantity of papers over time.

Given an heterogeneous social network, can we forecast its future? Can we predict who will start using a given hashtag on twitter? Can we leverage side information, such as who retweets or follows whom, to improve our membership forecasts? We present TensorCast, a novel method that forecasts time-evolving networks more accurately than current state of the art methods by incorporating multiple data sources in coupled tensors. TensorCast is (a) scalable, being linearithmic on the number of connections; (b) effective, achieving over 20% improved precision on top-1000 forecasts of community members; (c) general, being applicable to data sources with different structure. We run our method on multiple real-world networks, including DBLP and a Twitter temporal network with over 310 million non-zeros, where we predict the evolution of the activity of the use of political hashtags.

Exploiting resource reusability and low precision in neural networks is a promising approach to achieve energy efficient computational platforms. This research presents two generalizable approaches to reuse resources in feed-forward neural networks and demonstrated on extreme learning machines. In the first approach, coalescing, a single stack of neuronal units perform both feature extraction and classification tasks through shared resources. In the second approach, folding, the neurons in a high-dimension feedforward layer are folded to execute multiple-tasks. The folding technique can also be combined with low precision modules. The proposed design techniques are validated for a classification task on binary (Australian credit and Diabetes corpus) and multi-class (MNIST) dataset. The total power consumption is measured to be 3.65 mW on TSMC 65nm technology node, while yielding an accuracy of 91.7% for MNIST.

In the Area of Security, Intrusion Detection System (IDS) form an individual trailing and plays an essential role in information Security. As the usability of the internet among the users in a wide area is increasing day by day so as the importance of security and to keep the system aware of the malicious activities is also increasing. It has the following limitations on low detection rate, high false alarm rate and so on which is been indicated by the traditional Intrusion Detection System. A performance of the classifier is based on the necessity of the terms of its effectiveness, and it is also concerned with the number of features to be examined by the IDS should be improved. In this, J48 is been performed on the hybrid IDS and is applied using J48 Decision Tree algorithm, J48 Decision Tree is used for the feature selection and Naive Bayes Algorithm. Basically Intrusion detection systems (IDSs) is been used on the basis of two fundamental approaches first the recognition of anomalous activities as it generally occurs on the turns from usual or unusual behavior and second misuse detection by observing unauthorized “signatures” of those recognized malicious assaults and classification vulnerabilities. Anomaly or the anonymous (behavior-based) IDSs presume the difference of normal behavior beneath attacks and achieve abnormal activities evaluated and recognized with predefined system or user behavior reference model. The main focus of this survey is on WEKA (Waikato Environment for Knowledge Analysis) Tool and its various algorithms of classification used for detecting and analyzing the various intrusions. Lastly, In this survey, we lead to elaborate the mostly used dataset in information security research KDDCUP and NSL KDD and its various components.

Platforms for publishing research papers are increasing largely that contribute to big data as their volume is humongous and are unstandardized. Classification of this huge chunk of data is one of the biggest challenges in Information Retrieval. In this paper we discuss a scoring based unsupervised learning approach to extract relevant features and classify the research papers according to their content on a two class dataset. Feature extraction is carried out by analyzing the sections of a research paper and scoring them, followed by a two level hybrid classification technique based on title and conceptual summary using graph clustering. Promising experimental results are observed for data set using our research paper classification method. We present the experimental results of our proposed algorithm for feature extraction and clustering and compare the same with different approaches reported in literature.

Advancement of technology over the recent past has led to an increase in circulation of counterfeit notes in today's economy. To combat this issue, it is essential that there exists an efficient mechanism to detect fake banknotes. The main problem with current systems is the trade-off between speed and complexity. This paper proposes a system that can classify and subsequently verify Indian paper currency using fundamental image processing techniques. It uses the comparison between the input banknote and the calculated reference values for different parameters of original banknotes in a similar environment. This system maintains its simplicity while still having high accuracy of 100% for classification and 90% for validity verification.

In recent years, there has been an increasing need for the early detection of emerging research fronts. Research in this field usually employs citation networks, but this methodology does not address the citation lag problem. Text information is required to solve the time gap in citation networks because text information is available immediately when papers are published. However, text information has an inherent domain dependency problem. To address this, we introduce the “Dynamic Topic Model” (DTM). In a DTM, text information is represented in an abstract “topic” form and text information is captured as an increase or decrease in topics. We apply a DTM to the nanocarbon domain, which has experienced significant structural changes. We note that the choice of a suitable number of topics for the DTM requires further research. In this paper, we show that the proposed methodology, text information analysis with a DTM, can detect emerging research fronts earlier than the citation network technique.

Semantic links within text play an important role in understanding text. Cause-effect link is one of the basic semantic links that represents logic order between representation components of text. The importance of cause-effect link has been recognized and investigated in linguistics, but it has not been accurately measured to support computing applications. This paper is to investigate the role of cause-effect link within scientific paper. Research is conducted along two paths: (1) Human observations: Professionals find cause-effect links within a set of given papers, and then observe the number, the distribution and the keywords coverage of cause-effect links within each paper. The statistical results show that cause-effect links cover 76% keywords within paper on average. (2) Automatically discover more cause-effect links within a set of papers by developing a pattern-based algorithm. The automatically discovered cause-effect links validate the properties drawn from human observations. Experiments show that the algorithm can extract more than 80% of manually labeled links, and the automatically extracted links contain 75% of keywords within paper on average.

Research papers are often referred by researchers. But finding the desired or relevant research paper quickly and accurately is very difficult. As there are lot of research papers in given dataset and keeps enormously increasing. There is a need to avail automated processing approach for tackling such a huge volume of dataset to retrieve relevant papers accurately. This paper proposes a framework which works in two phases. In first phase, papers are stored in a dataset. Initially the Information Gain is calculated for finding features of each document in dataset. These features of document are used for classification. For classification of document, Naive Bayes approach is used. In second phase, relevant documents are retrieved based on search query. In this features of search query are extracted using Information Gain. The extracted features are used for retrieving relevant research papers. The extracted features of query and features of document stored in a dataset are compared using cosine similarity approach.

A large amount of information is continuously generated in intensive health care. An analysis of these data streams can supply valuable insights to improve the monitoring of the patients. The volume, frequency and complexity of data, which come unlabeled, make their analysis a challenging task. Machine learning (ML) techniques have been successfully employed for mining data streams to extract useful knowledge for health care monitoring. It includes the detection of changes in the behavior of sensors, failures on machines or systems, and data anomalies. Anomaly (or outlier) detection is a ML task that aims to find exceptions or abnormalities in a dataset. These exceptions, in a medical context, can represent a new disease pattern, an event to be further investigated, behavior changes or potential health complications. Despite of its analysis in data streams is a challenging task, temporal abstractions techniques should help due to they deal with the management and abstraction of time based data, offering high level of visualization of each data object in its context. The aim of this paper is to review recent research in anomaly detection and temporal abstractions and discuss the application of their combination to intensive care data streams.

Numerous security incidents caused by malwaresand hackers have recently utilized anti-forensic techniquesto bypass analysis and detection. It is critical to build aknowledge base that would help understand such anti-forensictechniques. In this paper, we present a forensic analysis methodto detect an anti-forensic technique which leverages timestampmanipulation in NTFS file system. Our approach analyzeshow timestamp manipulation occurs in NTFS file system andalso extracts some features to detect timestamp manipulationbehaviors. We also evaluate our approach with several usecases and describe how our approach helps detect timestampmanipulation behaviors.

Most works covering the topic of transfer learning propose an algorithm to solve a given domain adaptation problem, then test the algorithm using real-world datasets. A test with a real-world dataset represents a single transfer learning test condition, which partially measures an algorithm's performance. Previous research has placed little emphasis on developing a comprehensive and uniform test for transfer learning algorithms. With this in mind, a test framework is proposed, comprising of distortion profiles which define a comprehensive test suite. The unique contribution of this paper is the definition of a test framework that measures a more complete profile of a transfer learning algorithm's capability, facilitating the identification of relative poor and good performance areas. As a proof of concept, the test framework is used to test a homogeneous transfer learning algorithm. The test framework will be the basis for a number of future applications.

The Florida Public Hurricane Loss Model (FPHLM) is a public catastrophe model that integrates and regulates all key components, such as meteorology, engineering, and actuarial components, by following a certain workflow in the execution phase. The validation phase governed by an Automatic Data Validation (ADV) program simulates each modeled execution component with a large number of historical insurance data of specific hurricane events. The differences between the actual losses and the modeled losses of the insurance portfolios are evaluated to validate the model. The original validation process is time-consuming and error-prone when handling large data sets. This paper presents how the automated computer program efficiently and correctly incorporates the key components and produces useful reports for the validation purposes. By considering sixty-six combinations (i.e., the combination of one company and one hurricane) of the claim data, the FPHLM model adopts the largest set of portfolios comparing to the other four private models, which makes the validation process more challenging.

There is currently a big demand for automating big data analysis. In the data analysis field, data abstraction or summarization playes an important role in the extraction of generalized information from large scale data. We developped an artificial intelligence computer system with the aim of automating big data analysis and came up with a method that can abstract numerical type data (age, height, time, etc.). However, it could not abstract or summarize label type data (customer ID, product code, name, etc.). In the present work, we have developed a label abstraction method based on information entropy. Experiments using open real data showed that the proposed method achieved an extraction accuracy of 80% evaluated by f measure. We intended to apply the proposed method to our artificial intelligence and perform further evaluations.

This paper presents VEGAS - an online system that can illustrate the influence of one scientific paper on citation networks via the influence graph summarization and visualization. The system is built over an algorithm pipeline that maximizes the rate of influence flows in the final summarization. Both visualization and interaction designs are described with respect to a real usage scenario of the VEGAS system.

Data visualization traditionally is the most powerful tool for demonstration and analysis of scientific results and mathematical models in particular. In this paper we introduce the graphical framework for citation graph clustering. Furthermore, we discuss ways to detect factors responsible for scientific groups formation. Two datasets of scientific papers related to different fields were used in this work. Firstly we applied scientometric analysis to our data with the view to determine the most influential keywords. After that, we used two different ways for data clustering - graphic clustering method comprising N-body communication graph and a keyword-based hierarchical clustering. As a result of our studies we propose method for dynamic visualization of scientific papers clusters, built using open-access data.

Web usage mining refers to the process of discovering user access patterns from the log of website. Usually the web log contains unstructured, noisy and irrelevant data. To make this data suitable for pattern mining and pattern analysis it has to be passed through data preprocessing phase. Data preprocessing not only improves the quality of data but it also reduces the size of web log file. Data preprocessing involves several steps including data collection, data cleaning, session identification, user identification and path completion. This paper presents several data preprocessing techniques in order to prepare raw data suitable for mining and analysis tasks.

International audience; This paper presents our recent research on augmented maps where unprepared paper maps are used and Geographic Information System serves both as reference database and augmentation data. We use road intersections to register paper map and GIS so several paper maps of the same city whatever their styles and colors can be augmented using the same reference database. We present a new road extraction method to get better road intersection for a quick initialization and a faster method to register maps and GIS. Different maps are used to demonstrate the effectiveness of our method.

A system framework of test base management and intelligent test paper composition is investigated. Using ADO technology, table of the test paper in SQL server database can be managed. A software system of test database management and test paper generation has been developed. Based on the genetic and adaptive algorithm, a novel method is proposed for test paper composition. By the new method, the evolution speed can be improved, and local optimal solutions can be avoided the more efficiently. Analyzing of an example illuminates for the better performance of the new algorithm.

Traditional activity-centric process modeling languages treat data as simple black boxes acting as input or output for activities. Many alternate and emerging process modeling paradigms, such as case handling and artifact-centric process modeling, give data a more central role. This is achieved by introducing lifecycles and states for data objects, which is beneficial when modeling data-or knowledge-intensive processes. We assume that traditional activity-centric process modeling languages lack the capabilities to adequately capture the complexity of such processes. To verify this assumption we conducted an online interview among BPM experts. The results not only allow us to identify various profiles of persons modeling business processes, but also the problems that exist in contemporary modeling languages w.r.t. The modeling of business data. Overall, this preliminary empirical study confirms the necessity of data-awareness in process modeling notations in general.

Outlier detection is an important task in data mining, with applications ranging from intrusion detection to human gait analysis. With the growing need to analyze high speed data streams, the task of outlier detection becomes even more challenging as traditional outlier detection techniques can no longer assume that all the data can be stored for processing. While researchers mostly focus on detecting global outliers for data streams, detecting local outliers on streaming data has been neglected. This is an example of the utility problem in machine learning, where the machine learning algorithm needs to consider how the scarcity of a critical resource in the deployment environment affects the utility of any learned model. In this paper we focus on local outliers and propose an incremental solution assuming finite memory available. Our experimental results on a variety of data sets show that our solution is well suited to application environments with limited memory (e.g., wireless sensor networks) where the state of the system is changing.

Network science is essential in many civil and military applications due to its great help to complex systems with network-based structures. Our research is mainly about establishing network models to measure influence and impact. We establish two models to evaluate the significance of nodes for directed network(e.g:paper citation network) and undirected network(e.g:friend circle) respectively. Our models have relative accurate assessment on the influence of nodes in a specific network and are of great expansion ability and can be applied to different sets. These models are useful in solving practical problems.

This paper based on literature review introduces two powerful techniques-mapper and persistent homology in topological data analysis (TDA) to compute a topological abstraction of the printing model in 3D printing area. After leading in the theory of these methods and application of mapper for 3D printing, this paper also learns about the advantage of TDA than traditional methods for capturing the whole structure of targeted objects especially in high dimensional data. At last this paper learns about an example which combines mapper and persistent homology using dragon dataset to get evaluation of the quality of3D printed objects and reaches certain results. In conclusion, TDA in 3D printing is very potential. However, this field need to be further investigated.

We address domain adaptation in the context of clustering where we are given a set of unlabeled data, coming from several domains, and the goal is to group data into different categories regardless of the domain they come from. This is a challenging problem since we do not have any supervision unlike most adaptation scenarios studied earlier, and is very relevant in practical industry applications where labeled data often comes at a premium especially while deploying services that do not have a comparable predecessor. Our philosophy in addressing this problem draws motivation from the concept of dirty paper coding, a communications technique where the signal being transmitted through a noisy channel is encoded with priors on the possible noise patterns to assist reliable decoding of the signal at the receiver. We focus on image applications in this paper, where we encode priors on possible image domain shift factors such as viewpoint, lighting, blur and appearance variations and utilize geometric adaptation mechanisms to perform clustering. We illustrate the utility of our approach on standard datasets involving objects and faces, by obtaining around 18% improvement on average over existing approaches.

The motivation behind this paper is to enhance the reliability of in-vehicle navigation systems by predicting the duration of incidents that cause congestion. The main objective of this paper is to develop a methodology for predicting incident duration using broadcast incident data and evaluate the performance of k-NN and hazard-based duration models for predicting incident duration; both of the models are presented in this paper. An incident dataset from the BBC for the Greater London area is used to evaluate the accuracy of both models so that the results give a direct comparison between the models. The strengths and weaknesses of the models are discussed in the paper based on this analysis. Results show that both k-NN and hazard based models have the potential to provide accurate incident duration prediction. While k-NN based models provided marginally more accurate prediction than hazard-based models, the hazard-based duration models can provide additional information such as delay probabilities that can be used by advanced routing and navigation algorithms. Results also show that traffic information incident feeds, such as the tpegML feed from the BBC or TMC information, can be used as a potential data source for incident duration prediction in vehicle navigation systems.

This paper analyzes the time parameters of information, considers numerous characteristics including information value, information efficiency and so on, and proposes an accurate mathematical model for information timeliness. The mathematical model makes use of the active information-utilization-window (IUW) to analyze information value, effect of information efficiency on information timeliness, thus obtains the best IUW. With this window, the relationship among information timeliness, time-sensitive and time efficiency can be analyzed by using simulation. The method increasing information timeliness, proposed in this paper based on our analysis, is able to provide theoretical foundation for utilizing information of satellite network more reasonable.

Under the background of today's information age, micro blog obtains a rapid development. With the news on the micro blog updating, in order to avoid the users getting lost in the ocean of information, emotion analysis of the information becomes urgent and important. This paper based on the implementation of micro blog emotion mining of Bayesian classifier and SVM classification algorithm, making comparison through the analysis of the experimental results in processing speed and accuracy, has a reference value.

In this study, a novel method about building Deep Knowledge Graph for the Plant Insect Pest and Disease, namely DKG-PIPD, was proposed. Specifically, the semi-automatic extraction of semi-structured and unstructured knowledge was carried out on the basis of domain ontology, and the knowledge graph was stored in the third-party knowledge database according to the corpus characteristic of the plant insect pest and disease, to realize the visual display of entity interactive relationship and knowledge inference. Furthermore, DKG-PIPD performed joint extraction about the entity and the relationship in unstructured knowledge in a corpus tagging method that is suitable for domain data. In this way, the entity and the relationship were annotated synchronically, and the triplet can be obtained directly through label matching and label mapping, which not only effectively improved the annotation efficiency, but also solved the problem of one-versus-many overlapping relation extraction. In addition, DKG-PIPD used a novel end-to-end model to train and predict on the crawled dataset. The experimental contrast results with other classical benchmark methods demonstrated the effectiveness of the proposed method. Moreover, the related work in this paper first introduced the general architecture required for the building of knowledge graph, and then summarized its key points, that is, named entity recognition, entity relationship extraction and knowledge inference using deep learning are emphatically introduced. Finally, the improvement direction of this paper was also introduced in the discussion section.