Semantic Similarity Methods Research Articles

The dynamic effect of interactivity on customer engagement behavior through tie strength: Evidence from live streaming commerce platforms

• Interactivity has a nonlinear (inverted U-shaped) relationship with customer engagement. • Tie strength plays a completely mediating role of interactivity and customer engagement. • The various of customer ratings and tenure of memberships moderate the interactivity–tie strength relationship. • A semantic keyword similarity method is applied to customer reviews. Interactivity, which is a key characteristic of the live streaming commerce environment, fosters users’ active attitudes and behaviors in communications and transactions. However, the literature on live streaming commerce, is scarce, and few studies examine how interactivity influences customers’ non-transactional behaviors from a dynamic perspective. In this setting, based on the stimulus-organism-response (S-O-R) framework, we developed a research model using real-time data to investigate the dynamic effect of interactivity on customer engagement behavior through tie strength in live streaming commerce, which is a relatively new derivative of social commerce. This study developed a text mining method to quantify constructs using a large-scale sample of 3,500,445 online review texts. Our empirical study found that interactivity has a curvilinear relationship with customer engagement behavior. Besides, tie strength plays an intermediary role between interactivity and customer engagement behavior. It was further observed that both tenure of membership and popularity have an important moderating relationship between interactivity and tie strength. The study enriches the relationship marketing theory and live streaming commerce literature. Moreover, this study is one of the first studies to use real-time online data for live streaming commerce research.

INTEGRATING SEMANTIC SEARCH IN E-LEARNING TECHNOLOGIES: THE ELSE SYSTEM

The integration of semantic web methodologies and e-learning technologies is a challenge that has attracted a lot of attention for a decade. Given this, the purpose of this paper is the definition of a new e-learning semantic web methodology for the development of courses for health professionals in both distance and residential learning modes. ELSE is an ontology-based system which allows the construction of customized e-learning courses according to the needs and learning preferences of the user. It integrates semantic search methodologies and e-learning technologies. The underlying methodology relies on a reference domain ontology and teaching multimedial interactive modules, referred to as Reusable Learning Objects (RLOs), which are annotated according to the concepts of the ontology. The user can specify his/her training needs by selecting a set of concepts from the ontology, and the SemSim semantic search engine allows the identification of the set of RLOs that satisfy the user request at best, in efficient way. SemSim is a semantic similarity method which has been extensively experimented with and shows a higher correlation with human judgment with respect to the most relevant similarity methods defined in the literature. The set of RLOs is successively reorganized according to the learning preferences of the user. ELSE has been developed within a project of the CME (Continuing Medical Education) program - ECM for Italian participants - whose goal is the introduction of new methodologies and tools to keep updated health professionals and, in particular, medical specialists, in order to ensure effectiveness, safety, and efficiency of the national health service. ELSE has been tested and validated in the domain of osteoporosis, and the overall judgment about the system is very positive, both in terms of usability and effectiveness of customization. The system has been developed in cooperation with the ECM provider SPES S.c.p.A., accredited by the Italian Ministry of Health.

Measuring semantic similarity of documents with weighted cosine and fuzzy logic

Currently, the semantic analysis is used by different fields, such as information retrieval, the biomedical domain, and natural language processing. The primary focus of this research work is on using semantic methods, the cosine similarity algorithm, and fuzzy logic to improve the matching of documents. The algorithms were applied to plain texts in this case CVs (resumes) and job descriptions. Synsets of WordNet were used to enrich the semantic similarity methods such as the Wu-Palmer Similarity (WUP), Leacock-Chodorow similarity (LCH), and path similarity (hypernym/hyponym). Additionally, keyword extraction was used to create a postings list where keywords were weighted. The task of recruiting new personnel in the companies that publish job descriptions and reciprocally finding a company when workers publish their resumes is discussed in this research work. The creation of a new gold standard was required to achieve a comparison of the proposed methods. A web application was designed to match the documents manually, creating the new gold standard. Thereby the new gold standard confirming benefits of enriching the cosine algorithm semantically. Finally, the results were compared with the new gold standard to check the efficiency of the new methods proposed. The measures used for the analysis were precision, recall, and f-measure, concluding that the cosine similarity weighted semantically can be used to get better similarity scores.

Application of an Improved Synthetical Semantic Similarity Method in Water Knowledge Graph

In order to integrate different kinds of information in the water affairs field, this paper proposes an improved comprehensive semantic similarity algorithm and applies it to the construction of water affairs ontology. In this article, the algorithm is to give the concept instance similarity algorithm and concept definition similarity algorithm appropriate weight. First, we apply the three algorithms to the concept of the local ontology of smart water, and then calculate the similarity between the concepts to judge the similarity between the concepts to merge the smart water data, and finally compare the similarities of the three algorithms Degree value and calculation and comparison of three similarity variance values prove the effectiveness of the method. It can be seen from the experimental results that the variance of the comprehensive similarity calculation method is smaller than that of the other two methods, and it is easier to determine the similarity between two concepts.

IDSSIM: an lncRNA functional similarity calculation model based on an improved disease semantic similarity method

BackgroundIt has been widely accepted that long non-coding RNAs (lncRNAs) play important roles in the development and progression of human diseases. Many association prediction models have been proposed for predicting lncRNA functions and identifying potential lncRNA-disease associations. Nevertheless, among them, little effort has been attempted to measure lncRNA functional similarity, which is an essential part of association prediction models.ResultsIn this study, we presented an lncRNA functional similarity calculation model, IDSSIM for short, based on an improved disease semantic similarity method, highlight of which is the introduction of information content contribution factor into the semantic value calculation to take into account both the hierarchical structures of disease directed acyclic graphs and the disease specificities. IDSSIM and three state-of-the-art models, i.e., LNCSIM1, LNCSIM2, and ILNCSIM, were evaluated by applying their disease semantic similarity matrices and the lncRNA functional similarity matrices, as well as corresponding matrices of human lncRNA-disease associations coming from either lncRNADisease database or MNDR database, into an association prediction method WKNKN for lncRNA-disease association prediction. In addition, case studies of breast cancer and adenocarcinoma were also performed to validate the effectiveness of IDSSIM.ConclusionsResults demonstrated that in terms of ROC curves and AUC values, IDSSIM is superior to compared models, and can improve accuracy of disease semantic similarity effectively, leading to increase the association prediction ability of the IDSSIM-WKNKN model; in terms of case studies, most of potential disease-associated lncRNAs predicted by IDSSIM can be confirmed by databases and literatures, implying that IDSSIM can serve as a promising tool for predicting lncRNA functions, identifying potential lncRNA-disease associations, and pre-screening candidate lncRNAs to perform biological experiments. The IDSSIM code, all experimental data and prediction results are available online at https://github.com/CDMB-lab/IDSSIM.

Redundancy Detection Of Sentence Pairs In The Software Requirements Specification Documents With Semantic Approach

Software requirements specification document must complete, unambiguous and must correct because can lead to failure in software development. It must contain sentences not redundant. Meyer’s seven sins explain some errors in software requirements specification document, one of which is a Noise. Variant of the noise is redundancy. Redundancy is repeating the same information, but using different terms or phrases, thus giving the impression that new information. This research to analyze performance redundancy detection in sentences pair on software requirement specification documents using WordNet based semantic similarity methods. This study tries to develop a framework that has been proposed in previous studies in different framework that is to detected redundancy in sentences without fact extraction. The proposed approach and evaluation process uses Kappa values to determine whether reliable framework that can be used to detect redundancy in sentence pairs by using experiments in two scenarios to get good results. Our experimental result show the performance of proposed framework can used to help reduce redundancy when creating software requirements specification documents and may contributes to enhance the quality of document. Shown with good kappa values and interpretation of the Kappa value is substantial agreement. It means many agreements between experts and framework.

Evaluating cross-lingual textual similarity on dictionary alignment problem

Bilingual or even polylingual word embeddings created many possibilities for tasks involving multiple languages. While some tasks like cross-lingual information retrieval aim to satisfy users’ multilingual information needs, some enable transferring valuable information from resource-rich languages to resource-poor ones. In any case, it is important to build and evaluate methods that operate in a cross-lingual setting. In this paper, Wordnet definitions in 7 different languages are used to create a semantic textual similarity testbed to evaluate cross-lingual textual semantic similarity methods. A document alignment task is created to be used between Wordnet glosses of synsets in 7 different languages. Unsupervised textual similarity methods—Wasserstein distance, Sinkhorn distance and cosine similarity—are compared with a supervised Siamese deep learning model. The task is modeled both as a retrieval task and an alignment task to investigate the hubness of the semantic similarity functions. Our findings indicate that considering the problem as a retrieval and alignment problem has a detrimental effect on the results. Furthermore, we show that cross-lingual textual semantic similarity can be used as an automated Wordnet construction method.

Protein2vec: Predicting Protein-Protein Interactions Based on LSTM.

The semantic similarity of gene ontology (GO) terms is widely used to predict protein-protein interactions (PPIs). The traditional semantic similarity measures are based mainly on manually crafted features, which may ignore some important hidden information of the gene ontology. Moreover, those methods usually obtain the similarity between proteins from similarity between GO terms by some simple statistical rules, such as MAX and BMA (best-match average), oversimplifying the possible complex relationship between the proteins and the GO terms annotated with them. To overcome the two deficiencies, we propose a new method named protein2vec, which characterizes a protein with a vector based on the GO terms annotated to it and combines the information of both the GO and known PPIs. We firstly try to apply the network embedding algorithm on the GO network to generate feature vectors for each GO term. Then, Long Short-Time Memory (LSTM) encodes the feature vectors of the GO terms annotated with a protein into another vector (called protein vector). Finally, two protein vectors are forwarded into a feedforward neural network to predict the interaction between the two corresponding proteins. The experimental results show that protein2vec outperforms almost all commonly used traditional semantic similarity methods.

SimiT: A Text Similarity Method Using Lexicon and Dependency Representations

Semantic textual similarity methods are becoming increasingly crucial in text mining research areas such as text retrieval and summarization. Existing methods of text similarity have often been computed by their shallow or syntactic representation rather than considering their semantic content and meanings. This paper focuses mainly on computing the similarity between sentences without a supervised learning approach, only considering their word-level coherence which is calculated by a hybrid method of dependency parser and lexicon embeddings. Hence, we concentrate on structural similarity between text pairs by regarding their dependency parser embeddings. Our hybrid method also pays attention to the semantic information of words implied in the sentences. In the evaluation, we compare our method with the state-of-the-art semantic similarity measures in a well-known dataset. Our method outperforms most of the studies in the literature and the overall performance achieves better results when combining the similarity scores of both embedding models.

Visibility Tagging of nouns in Text-to-Scene Conversion

The recognition and extraction of visual text information is the basis of the realization of Text-to-Scene Conversion, while nouns are the main expression form of entity, event, orientation, action and other information in text, but not all nouns are suitable for visualization, so how to recognize the visibility of nouns is the key problem of Text-to-Scene Conversion. Based on the definition and classification of noun visualization, this paper establishes a keyword visibility dictionary. Furthermore, in view of the static limitations of the dictionary, a dynamic expansion method based on semantic similarity is studied, and an optimization method for the visibility annotation of nouns is proposed in combination with unsupervised clustering. The experimental results show that the proposed method is effective and feasible for nominal visibility annotation of common texts, and the accuracy of the optimized method is 16% higher than that of the simple semantic similarity method.

An approach for measuring semantic similarity between Wikipedia concepts using multiple inheritances

Wikipedia provides a huge collaboratively made semi-structured taxonomy called Wikipedia category graph (WCG), which can be utilized as a Knowledge Graph (KG) to measure the semantic similarity (SS) between Wikipedia concepts. Previously, several Most Informative Common Ancestor-based (MICA-based) SS methods have been proposed by intrinsically manipulating the taxonomic structure of WCG. However, some basic structural issues in WCG such as huge size, branching factor and multiple inheritance relations hamper the applicability of traditional MICA-based and multiple inheritance-based approaches in it. Therefore, in this paper, we propose a solution to handle these structural issues and present a new multiple inheritance-based SS approach, called Neighborhood Ancestor Semantic Contribution (NASC). In this approach, firstly, we define the neighborhood of a category (a taxonomic concept in WCG) to define its semantic space. Secondly, we describe the semantic value of a category by aggregating the intrinsic IC-based semantic contribution weights of its semantically relevant multiple ancestors. Thirdly, based on our approach, we propose six different methods to compute the SS between Wikipedia concepts. Finally, we evaluate our methods on gold standard word similarity benchmarks for English, German, Spanish and French languages. The experimental evaluation demonstrates that the proposed NASC-based methods remarkably outperform traditional MICA-based and multiple inheritance-based approaches.

An Efficient Approach for Measuring Semantic Similarity Combining WordNet and Wikipedia

The measurement of semantic similarity between concepts is an important research topic in natural language processing. In the past, several approaches for measuring the semantic similarity between concepts have been proposed based on WordNet or Wikipedia. However, improvements in the measurement accuracy of most methods have led to a dramatic increase in time complexity, and the existing methods do not effectively integrate WordNet and Wikipedia. In this paper, we focus on designing an efficient semantic similarity method based on WordNet and Wikipedia. To improve the accuracy of WordNet edge-based measures, we propose an edge weight model for combining edge and density information, which assigns a weight to each edge adaptively based on the number of direct hyponyms of the subsumer. Second, to improve the computational efficiencies of the existing Wikipedia link vector-based measures, we propose a new Wikipedia link feature-based semantic similarity method that converts Wikipedia links into semantic knowledge and replaces the TF-IDF statistical weight model in the existing measures. In addition, we propose two new word disambiguation strategies to further improve the accuracy of Wikipedia link-based measures. Finally, to fully exploit the advantages of WordNet and Wikipedia, we propose two new aggregation schemas for combining WordNet “ is-a ” semantics and Wikipedia link semantics to replace the current aggregation schemas that combine WordNet “ is-a ” semantics with category semantics in Wikipedia. The experimental results show that our aggregation models are outstanding in terms of accuracy, efficiency and word coverage compared to state-of-the-art similarity measures.

Text Classification Based on LDA and Semantic Analysis

The quality of text features directly affects the text classification effect, in order to get the text features which have the high contribution to the text classification in class, this paper proposes a text classification method based on LDA model and category semantic similarity method. The method selects text document topic features by the LDA model and calculates the semantic similarity between these features and categories combined with the word vector model. According to the size of similarity, the weight of the text feature is obtained, and the text classification feature selection and text classification are realized. Finally, the feasibility, validity and correctness of the algorithm are verified by experiments.

The K-Means Clustering Algorithm With Semantic Similarity To Estimate The Cost of Hospitalization

The cost of hospitalization from a patient can be estimated by performing a cluster of patient. One of the algorithms that is widely used for clustering is K-means. K-means algorithm, based on distance still has weaknesses in terms of measuring the proximity of meaning or semantics between data. To overcome this problem, semantic similarity can be used to measure the similarity between objects in clustering, so that, semantic proximity can be calculated. This study aims to conduct clustering of patient data by paying attention to the similarity of the patient’s disease. ICD code is used as a guide in determining a patient’s disease. The K-means method is combined with semantic similarity to measure the proximity of the patient’s ICD code. The method used to measure the semantic similarity between data, in this study, is the semantic similarity of Girardi, Leacock & Chodorow, Rada, and Jaccard Similarity. Cluster quality measurement uses the silhouette coefficient method. Based on the experimental results, the method of measuring semantic similarity data is capable to produce better quality clustering results than without semantic similarity. The best accuracy is 91.78% for the three semantic similarity methods, whereas without semantic similarity the best accuracy is 84.93%.

Generating Requirement Dependency Graph Based on Class Dependency

Requirements association depicts inter-relation between two or more requirements within a software project. It provides necessary information for developers during decision-making processes, such as change management, development milestones, bug prediction, cost estimation, and work breakdown structure generation. Modeling association between requirements became a focus of software requirements researchers. Previous studies indicate that requirements association was pre-defined by requirements engineer based on their expert judgments. The judgments require knowledge on requirements and their class realizations. This paper introduces a method to generate a mapping between a set of requirement statements and a set of classes of a given project that realized the respected requirements. The method also generates associations among requirements based on information on associations between classes and the class-requirement mapping. The method utilizes element of relational information resided in a class diagram of respected project. A semantic similarity method was used to define the requirements with their realization classes. A class is considered realizing a requirement if and only if their semantic similarity is higher than a certain threshold. A set of experimentation on four different projects was conducted. The result of the approach was compared with the output produced by human annotators using kappa statistics. The approach is considered as having a fair agreement level (i.e. with kappa value 0.37) with the human annotators to identify and model requirement associations.

Perbandingan Metode Collaborative Filtering dan Hybrid Semantic Similarity

Recommender system is a component which has been developed for online commerce purposes. In this issue, one of the popular methods that has been widely used is collaborative filtering. However, this method has some drawbacks and needs to be improved. Therefore, in this research a combination of Collaborative Filtering (CF) and semantic similarity method has been compare with original CF, and the result expected reducing some deficiencies on the original collaborative filtering method. Based on the performance tests, the results conclude that the combination can reduce some weaknesses on the original collaborative filtering, especially on the cold-start item and sparsity issue.

Semantic textual similarity between sentences using bilingual word semantics

Semantic textual similarity between sentences is indispensable for many information retrieval tasks. Traditional lexical similarity measures cannot compute the similarity beyond a trivial level. Moreover, they only can capture the textual similarity, but not semantic. In this paper, we propose a method for semantic textual similarity that leverages bilingual word-level semantics to compute the semantic similarity between sentences. To capture word-level semantics, we employ distribute representation of words in two different languages. The similarity function based on the concept-to-concept relationship corresponding to the words is also utilized for the same purpose. Multiple new semantic similarity measures are introduced based on word-embedding models trained on two different corpora in two different languages. Apart from these, another new semantic similarity measure is also introduced using the word sense comparison. The similarity score between the sentences is then computed by applying a linear ranking approach to all proposed measures with their importance score estimated employing a supervised feature selection technique. We conducted experiments on the SemEval Semantic Textual Similarity (STS-2017) test collections. The experimental results demonstrated that our method is effective for measuring semantic textual similarity and outperforms some known related methods.

Answer selection using Word Alignment based on Part of Speech Tagging in community question answering

This paper explain about answer selection using word alignment based on POS tagging in Community Question-Answering (CQA). This online community allowed the user to ask and reply related to the question problems which has no restrictions. This causes inappropriate comments with the question problems proposed before. To solve these problems, combining lexical and semantic features has been developed with result conclude that the approach more adequate for similarity task rather than question answering. According to the previous research, there is several problems that can be enhanced. First, vector representation counts exactly matched words, so it does not effective to cover other words that have relatedness between two pairing words. Second, noun overlap for similarity measure in pairing words can’t define that the two words are similar. So, it must be define that the pairing POS tag is the same meaning or relatedness. In this study, unsupervised lexical and semantic similarity method employed with different approach from previous method in verbatim and contextual similarities. The data was taken from SemEval 2017 competition which focus on Question-Answer Similarity task. The experiment result for precision (Mean Average Precision) score shows the improvement from 0.674 to 0.6845, 1.03 % higher than previous research in CQA. This improvement comes from lexical similarity, which is not just from noun pattern but also taken from verb pattern. Furthermore, semantic similarity has an important role in determining which words that have same pattern and meaning to define relevancy between them.

Novel phenotype–disease matching tool for rare genetic diseases

PurposeTo improve the accuracy of matching rare genetic diseases based on patient’s phenotypes. MethodsWe introduce new methods to prioritize diagnosis of genetic diseases based on integrated semantic similarity (method 1) and ontological overlap (method 2) between the phenotypes expressed by a patient and phenotypes annotated to known diseases. ResultsWe evaluated the performance of our methods by two sets of simulated data and one set of patient’s data derived from electronic health records. We demonstrated that the two methods achieved significantly improved performance compared with previous methods in correctly prioritizing candidate diseases in all of the three sets. Our methods are freely available as a web application (https://gddp.research.cchmc.org/) to aid diagnosis of genetic diseases. ConclusionOur methods can capture the diagnostic information embedded in the phenotype ontology, consider all phenotypes exhibited by a patient, and are more robust than the existing methods when phenotypes are incorrectly or imprecisely specified. These methods can assist the diagnosis of rare genetic diseases and help the interpretation of the results of DNA tests.

Measuring Customer Agility from Online Reviews Using Big Data Text Analytics

Large volumes of product reviews generated by online users have important strategic value for product development. Prior studies often focus on the influence of reviews on customers‘ purchasing decisions through the word-of-mouth effect. However, little is known about how product developers respond to these reviews. This study adopts a big data analytical approach to investigate the impact of online customer reviews on customer agility and subsequently product performance. We develop a singular value decomposition-based semantic keyword similarity method to quantify customer agility using large-scale customer review texts and product release notes. Using a mobile app data set with over 3 million online reviews, our empirical study finds that review volume has a curvilinear relationship with customer agility. Furthermore, customer agility has a curvilinear relationship with product performance. Our study contributes to innovation literature by demonstrating the influence of firms‘ capability of utilizing online customer reviews and its impact on product performance. It also helps reconcile inconsistencies found in literature regarding the relationships among the three constructs.