Semantic Mining Research Articles

A Hybrid Fuzzy System via Topic Model for Recommending Highlight Topics of CQA in Developer Communities

Question-answering (QA) websites supply a quickly growing source of useful information in numerous areas. These platforms present novel opportunities for online users to supply solutions, they also pose numerous challenges with the ever-growing size of the QA community. QA sites supply platforms for users to cooperate in the form of asking questions or giving answers. Stack Overflow is a massive source of information for both industry and academic practitioners, and its analysis can supply useful insights. Topic modeling of Stack Overflow is very beneficial for pattern discovery and behavior analysis in programming knowledge. In this paper, we propose a framework based on the Latent Dirichlet Allocation (LDA) algorithm and fuzzy rules for question topic mining and recommending highlight latent topics in a community question-answering (CQA) forum of developer community. We consider a real dataset and use 170,091 programmer questions in the R language forum from the Stack Overflow website. Our result shows that LDA topic models via novel fuzzy rules can play an effective role for extracting meaningful concepts and semantic mining in question-answering forums in developer communities.

Research of Semantic Considered Tree Mining Method for an Intelligent Knowledge-Services Platform

Research of Semantic Considered Tree Mining Method for an Intelligent Knowledge-Services Platform

DE-Ada*: A novel model for breast mass classification using cross-modal pathological semantic mining and organic integration of multi-feature fusions

Computer-aided breast mass classification is an effective and widely used technology to assist pathologists in formulating clinical diagnoses and improving working efficiencies. Existing studies usually use a single image feature to perform breast mass classification. Herein, we propose a simple, yet effective, model called the DE-Ada*, which is an organic integration of multi-feature fusions, for breast mass classification. Firstly, we extract a set of complementary features, namely, scale-invariant feature transform (SIFT), GIST, histogram of oriented gradient (HOG), local binary pattern (LBP), residual network (ResNet), densely connected convolutional networks (DenseNet), and visual geometry group (VGG), to characterize mammograms from diverse perspectives. We attempt to mine the cross-modal pathological semantics among these features and complete their early fusion. The dynamic weight of any feature or cross-modal pathological semantics is computed and utilized to complete mid-level feature fusion. Finally, we design two voting-based ensemble learning strategies to implement late feature fusion. Our experiments demonstrate that the DE-Ada* model outperforms baselines on two well-known mammographic datasets. Our model encourages the use of cross-modal pathological semantics to deal with the overfitting problem.

Nonindependent Session Recommendation Based on Ordinary Differential Equation

The recommendation method based on user sessions is mainly to model sessions as sequences in the assumption that user behaviors are independent and identically distributed, and then to use deep semantic information mining through Deep Neural Networks. Nevertheless, user behaviors may be a nonindependent intention at irregular points in time. For example, users may buy painkillers, books, or clothes for different reasons at different times. However, this has not been taken seriously in previous studies. Therefore, we propose a session recommendation method based on Neural Differential Equations in an attempt to predict user behavior forward or backward from any point in time. We used Ordinary Differential Equations to train the Graph Neural Network and could predict forward or backward at any point in time to model the user's nonindependent sessions. We tested for four real datasets and found that our model achieved the expected results and was superior to the existing session-based recommendations.

Discovering City Profile Based on Tag Semantic Mining

Discovering City Profile Based on Tag Semantic Mining

A comprehensive survey on trajectory-based location prediction

Recent explorative growth in telecommunication and telepathy technology has flooded the market with location-based data, which paves the way for location-aware prediction services. These applications have vast domain influence in route navigation, recommendation system, traffic-congestion control, ecological study, climatological forecast, and many more. Research efforts are spent on the put-forward overall picture of location prediction through trajectory data. This survey offers an extensive overview of location prediction enveloping basic definitions and concepts, data sources, approaches, and applications. Moreover, Spatial–Temporal pattern-based prediction models are discussed, highlighting the advantages and disadvantages of each. Sequential, periodic, and frequent pattern mining advances are noted. This paper presents a recent deep learning methodology for extracting features of a large trajectory. Distributive big data models using Hadoop and MapReduce frameworks are recorded. Location prediction using the social media platform is mentioned. Content-based and semantic mining models are studied. Tables and diagrams are displayed to provide at a glance view to facilitate smooth understanding. Furthermore, application and challenges are addressed related to the next location prediction. The overall conclusion of the survey and future directions are also listed.

Semantic process mining: A conceptual application of main tools, framework and model analysis

Semantics has been a major challenge when applying the process mining (PM) technique to real-time business processes. The several theoretical and practical efforts to bridge the semantic gap has spanned the advanced notion of the semantic-based process mining (SPM). Fundamentally, the SPM devotes its methods to the idea of making use of existing (semantic) technologies to support the analysis of PM techniques. In principle, the semantic-based process mining method is applied through the acquisition and representation of abstract knowledge about the domain processes in question. To this effect, this paper demonstrates how the semantic concepts and process modelling (reasoning) methods are used to improve the outcomes of PM techniques from the syntactic to a more conceptual level. To do this, the study proposes an SPM-based framework that shows to be intelligent with a high level of semantic reasoning aptitudes. Technically, this paper introduces a process mining approach that uses information (semantics) about different activities that can be found in any given process to make inferences and generate rules or patterns through the method for annotation, semantic reasoning, and conceptual assertions. In turn, the method is theoretically applied to enrich the informative values of the resultant models. Also, the study conducts and systematically reviews the current tools and methods that are used to support the outcomes of the process mining as well as evaluates the results of the different methods to determine the levels of impact and its implications for process mining.

Knowledge-based semantic discretisation using data mining techniques

Certain data mining techniques can be applied only on discretised data. Various studies show significant improvement in certain data mining techniques, when applied on discretised data rather than continuous data. Discretisation methods based on statistical techniques are inadequate in capturing and exploiting the knowledge inherent in data and context of study. To overcome this limitation, we propose a novel knowledge-based semantic discretisation method using data mining techniques, in which discretisation is done based on semantic data. Semantic data is domain knowledge inherent in the data and context of the study. Unlike semantic data mining, no explicit ontology associated with the data for semantic discretisation. Therefore, it's a challenging task to identify, capture, interpret and exploit the semantic data for semantic discretisation. This study presents the novel concept of semantic discretisation. The effectiveness of the proposed methodology is shown by applying it on Pima Indian diabetes dataset.

Analysis of the Douban online review of the MCU: based on LDA topic model

With the rapid development of the Internet and the movie industry, online reviews have shown explosive growth. Online review data contains a lot of valuable information, through text mining techniques such as the LDA topic model to analyze this information, to help the movie creators to understand the public’s viewing needs, movie production team to reflect on the shortcomings in the production process, in order to promote the development of the movie industry to provide a reference. In this paper, the Douban online review data of the MCU are used as the research object, and the Chinese text is analyzed based on the LDA topic model. Firstly, data collection and preprocessing are carried out with various software. Then, the word cloud is used to visualize the core information in the online review data, and then the LDA topic model is used for a deeper semantic mining. The research results show that the rich and varied plot design, the audiovisual feast brought by advanced digital technology, the clever connection of bonus scene, the unique characters and actors and a super-hero’s commercial movie mode are closely related to the success of the MCU.

Data mining model based on user reviews and star ratings

With the rapid development of e-commerce, the research on sentiment analysis of online reviews has been paid more and more attention. This paper presents an Aspect-Level Sentiment Analysis Method based on long short-term memory (LSTM) and boot-strapping, which performs semantic mining and prediction on time-based data patterns and data combinations of text, star rating and helpful votes. A high prediction accuracy rate is obtained in the open data set. Compared with the traditional methods, which single analysis comment or evaluation, merchants can gain a deeper understanding of user feedback from sentiment analysis.

A Status Property Classifier of Social Media User's Personality for Customer-Oriented Intelligent Marketing Systems

Enterprises need to obtain information about not only specific customer preferences, but also, more importantly, customers' psychological characteristics that significantly influence their consumption behaviors and response to intelligent-based marketing activities. If enterprises want to implement more precise intelligent selling activities for customers, customers' personality information will serve as a highly valued reference. The automatic detection method proposed in this study is based on techniques such as text semantic mining and machine learning to conduct personality type prediction on the target by collecting and analyzing the target's social media data. In the test, 5,858 statuses were obtained, 815 of which were labeled, with 122 effective tags. In general, when n = 5, the labeling rate can reach 60-80%. The status property classifier (SPC) proposed in this study can predict the personality type (PT) of the user publishing the status set with a high degree of accuracy by conducting text semantic mining on the status set.

A Domain Expertise and Word-Embedding Geometric Projection Based Semantic Mining Framework for Measuring the Soft Power of Social Entities

Social science research (communication studies in particular) often involves measuring the soft power of social entities from the related big data of media texts. This work proposes a soft power measurement framework that integrates the domain-expertise-driven conceptualization of the traditional social science paradigm and the data-driven operationalization of the computational social science paradigm. Based on the geometric projections of word vectors from the social entity keywords to the soft power dimension keywords, the weights of social entity keywords in the full-sample media texts are projected to the corresponding soft power dimensions in the high-dimensional word-embedding semantic space and accumulated to the total and dimensional values of the soft power of social entities respectively. We demonstrate the applicability, compatibility, computational complexity, validity, adaptability, and scalability of the proposed framework by showing a standard instantiation (measuring the national image carrying capacity of Chinese elite private entrepreneurs from the full-sample WeChat articles, each of which has more than 100,000 unique views) and a variational instantiation (measuring the city brand influence of global cities from the big data of the Global News Knowledge Graph in the Global Database of Events, Language, and Tone (GDELT) of Google) of the proposed framework.

Multidimensional Extra Evidence Mining for Image Sentiment Analysis

Image sentiment analysis is a hot research topic in the field of computer vision. However, two key issues need to be addressed. First, high-quality training samples are scarce. There are numerous ambiguous images in the original datasets owing to diverse subjective cognitions from different annotators. Second, the cross-modal sentimental semantics among heterogeneous image features has not been fully explored. To alleviate these problems, we propose a novel model called multidimensional extra evidence mining (ME 2 M) for image sentiment analysis, it involves sample-refinement and cross-modal sentimental semantics mining. A new soft voting-based sample-refinement strategy is designed to address the former problem, whereas the state-of-the-art discriminant correlation analysis (DCA) model is used to completely mine the cross-modal sentimental semantics among diverse image features. Image sentiment analysis is conducted based on the cross-modal sentimental semantics and a general classifier. The experimental results verify that the ME 2 M model is effective and robust and that it outperforms the most competitive baselines on two well-known datasets. Furthermore, it is versatile owing to its flexible structure.

WITHDRAWN: Video structure syntax mining based on Hidden Markov Model in sports video

Abstract On the basis of shot segmentation, an improved Rate Sensitive Competitive Learning algorithm (RSCLA) is proposed for unsupervised shot clustering based on NBA basketball video, which converts video stream data into symbol sequence. Aiming at the characteristics of sequence correlation, time correlation and no clear transaction concept in Video Association rules, this paper improved the traditional Apriori algorithm and proposed a video Association Rules Mining Algorithm based on time-based window computing support. The periodic or semi-periodic structural grammar patterns in video are explored with frequent sets of association rules to guide the establishment of video structure semantic mining model. Aiming at the limitation of string matching, this paper proposes a HMM-based pattern mining method to analyze high-level video events, identify and locate free throw events in basketball videos. The experimental results show that the four-state HMM is the inner free throw mode of basketball video and achieves good performance.

A text semantic topic discovery method based on the conditional co-occurrence degree

The topic discovery method, as an effective tool for semantic mining and a key means to extract new features from original text, plays an important role in the field of text mining and knowledge discovery. To solve the problems encountered in traditional topic models, such as the loss of semantic information and the ambiguity of topic concepts, as well as the crossover and coverage among topics, we propose a semantic topic discovery method based on the conditional co-occurrence degree (CCOD_STDM). First, every document is split into multiple subdocuments according to the semantic structure of the document and the independence decision rules. Second, combinatorial words with strong semantic relevance are extracted based on the conditional co-occurrence degree within the subdocuments. Based on these combinatorial words, new subdocuments are formed by feature expansion and content reconstruction. Third, “topic-word” distributions and “document-topic” distributions of new subdocuments are obtained by topic modeling with Gibbs sampling. Finally, “document-topic” distributions of the original documents are obtained by merging new subdocuments’ “document-topic” distributions with specific strategies. The numerical experiments are compared with six topic models and two evaluation methods on seven kinds of public corpora, and the experimental results verify the superiority of CCOD_STDM and its efficiency in topic discovery. More importantly, a case study illustrates that the combinatorial words can effectively avoid the polysemy problem and can facilitate the condensation and summary of topics.

Social learning analytics for determining learning styles in a smart classroom

ABSTRACT Social Learning Analytics (SLA) seeks to obtain hidden information in large amounts of data, usually of an educational nature. SLA focuses mainly on the analysis of social networks (Social Network Analysis, SNA) and the Web, to discover patterns of interaction and behavior of educational social actors. This paper incorporates the SLA in a smart classroom. Specifically, this paper proposes to determine the learning styles of the students in a smart classroom using SLA. In this proposal is analyzed external data from the web and social networks to build knowledge models about the students, in order to improve the learning processes that occur in the smart classrooms. In general, these SLA tasks will be organized in autonomous cycles, in order to integrate them with each other. The autonomic cycle will automate the execution of those tasks and the generation of knowledge models, in such a way to permanently monitor the learning process, observing it, analyzing it and determining the student learning styles. For the development of the SLA tasks, we will use concepts from the Semantic Mining, Text Mining, Data Mining, among other domains. Finally, we experiment in a test scenario, with results very interesting.

Building Intelligent Semantic Educational System (ISES) Based on Ontology and Semantic Web Mining.

Building Intelligent Semantic Educational System (ISES) Based on Ontology and Semantic Web Mining.

Systematic analysis and identification of unexpected interactions from the neuroprotein drug interactome in hydrocephalus pharmacological intervention.

Hydrocephalus is a neurological condition caused by an abnormal accumulation of cerebrospinal fluid; pharmacological intervention of the disease has been found to elicit a variety of adverse drug reactions (ADRs) in central nervous system (CNS) by unexpectedly targeting certain functional neuroproteins. Here, a systematic neuroprotein drug interactome (SNDI) is created for 11 hydrocephalus drugs/metabolites plus 20 control drugs across 518 druggable pockets on the surface of 472 CNS neuroproteins via a large-scale molecular docking approach. Heuristic clustering analysis of the SNDI profile divides the 31 investigated drug ligands into a distinct panel and a background panel; the former consists of two hydrocephalus drugs (Furosemide and Triamterene) and their respective metabolites (Furosemide glucuronide and Hydroxytriamterene) that are inferred to have generally high affinity towards the whole array of neuroprotein pockets. A total of 13 neuroproteins are enriched in gene ontology semantic mining as putative unexpected targets of the distinct panel, and their intermolecular interactions with hydrocephalus drugs/metabolites are investigated in detail using dynamics simulation and energetics analysis. We also perform kinase assay and viability test to substantiate the interactome analysis. It is found that the Furosemide and Triamterene have significant cytotoxic effects on normal human astrocytes, in which the Triamterene can inhibit the neurokinase ROCK2, a representative of putative unexpected targets, with a high activity, which is comparable with the sophisticated ROCK2 inhibitor Fasudil.

CBSSD: community-based semantic subgroup discovery

Modern data mining algorithms frequently need to address the task of learning from heterogeneous data, including various sources of background knowledge. A data mining task where ontologies are used as background knowledge in data analysis is referred to as semantic data mining. A specific semantic data mining task is semantic subgroup discovery: a rule learning approach enabling ontology terms to be used in subgroup descriptions learned from class labeled data. This paper presents Community-Based Semantic Subgroup Discovery (CBSSD), a novel approach that advances ontology-based subgroup identification by exploiting the structural properties of induced complex networks related to the studied phenomenon. Following the idea of multi-view learning, using different sources of information to obtain better models, the CBSSD approach can leverage different types of nodes of the induced complex network, simultaneously using information from multiple levels of a biological system. The approach was tested on ten data sets consisting of genes related to complex diseases, as well as core metabolic processes. The experimental results demonstrate that the CBSSD approach is scalable, applicable to large complex networks, and that it can be used to identify significant combinations of terms, which can not be uncovered by contemporary term enrichment analysis approaches.

A novel approach on Particle Agent Swarm Optimization (PASO) in semantic mining for web page recommender system of multimedia data: a health care perspective

Recent decades have seen huge amounts of information collected in clinical databases for mining patients’ health states from multimedia data. Since on the web diverse type of web recommendation is made obtainable toward user every day with the purpose of consists of Image, Video, Audio, query suggestion and web page. Therefore, multimedia data accessible designed for patient-oriented decision making has improved significantly however is often scattered across various sites. In this perspective, Web page recommender systems with multimedia data might provide patients with further laymen-friendly information helping toward enhanced understand their health status as represented by their record. In this research work, Web Page health recommender systems are introduced via the use of certain agents in order to provide extremely appropriate web pages for patients. The main feature of Particle Agent Swarm Optimization (PASO) is that the creation of the algorithm is denoted by a set of Particle agents who cooperate in attaining the objective of the task under consideration. In the research method, two kinds of agents are presented: web user particle agent and semantic particle agent. PASO Based Web Page Recommendation (PASO-WPR) system is an intermediate program (or a particle agent) containing a user interface, which wisely produces a collection of info that suits an individual’s requirements. PASO-WPR is carried out dependent upon incorporating semantic info with data mining techniques on the web usage data as well as clustering of pages dependent upon similarity in their semantics. As the Web pages with multimedia files are viewed as ontology individuals, the pattern of patients’ navigation are like instances of ontology rather than the uniform resource locators, and with the help of semantic similarity, page clustering is carried out. For producing web page recommendations to users, the outcome is utilized. The recommender engine concentrates on the semantic info and as well exploits a particle agent to reform the outcomes of web pages recommendation. Consequently, the system response time is enhanced and as a result, creating the framework scalable. The outcomes recommend that the PASO-WPR system is improved in identifying the web page that a user is about to request while matched up other approaches. The outcome proves that the presented PASO-WPR system is carried out well in regard to the accuracy measures aspects such as accuracy, coverage and M-Metric are identified to contain greater values compared to the previous item based collaborative filtering recommendation systems.