885 Scientific Paper Argumentation Ontology and Annotation Experiment Hot! Wang Xiaoguang, Zhou Huimin, Song Ningyuan DOI: 10.3772/j.issn.1000-0135.2020.09.001 Argumentation structure, an integral part of scientific papers, is an important kind of tacit knowledge that relates the claims made by scientific papers to their argumentation process. Normative description and accurate representation of scientific papers' argumentation structure is of great significance to semantic enhancement of scientific papers and document-based knowledge discovery. Based on argumentation theory, this study constructs a new Scientific Paper Argumentation Ontology (SAO) by reusing the existing document component ontology. SAO includes 7 core classes, 13 extended classes, and 15 attribute relationships. To evaluate its usability, 40 scientific papers from the fields of library and information science and biomedical science were selected for semantic annotation experiments. The statistical results showed that SAO has strong expression ability, and that although there are some similarities between the argumentation structure of scientific papers in different fields, their argumentation modes differ. 2020 Vol. 39 (9): 885-895 [Abstract] ( 246 ) HTML (115 KB)  PDF (1714 KB)  ( 743 )

896 Research on Influence Evaluation of Humanities and Social Sciences Academic Monographs from the Perspective of Altmetrics: Comparative Analysis Based on BkCI, Amazon, and Goodreads Hot! Li Jiangbo, Zhang Liang, Jiang Chunlin DOI: 10.3772/j.issn.1000-0135.2020.09.002 Altmetrics has developed rapidly since it was introduced, significantly broadening the research scope of bibliometrics evaluation. Currently, the vast majority of altmetrics research focuses on evaluating the academic achievements of papers, and it seems that not enough attention is paid to the evaluation of academic monographs or other types of academic achievements. However, the monograph is an important form of academic achievement in Social Sciences and Humanities. This study focuses on the evaluation of academic achievements of the monograph. By standardizing the publishing time, we calculate the daily average citation of monographs in the database. This study also uses a recurrent neural network method to classify the online reviews of monographs. It uses an emotion dictionary to analyze the review text with fine-grained emotion and gets the sentiment analysis indicator. The results show that some academic books are cited less in the BkCI database, but their daily average citation is higher because these academic monographs are published later, and there is not enough time to accumulate citations. This shows that the low citation of monographs does not mean that their academic influence should be low. It is problematic to evaluate academic monographs’ influence only by using the citation in the citation index database. Another result shows little correlation between Altmetrics indicators and citation indicators, especially between average sentiment score indicator and citation indicator. Low correlation means that the Altmetric indicators from online reviews do not have the feasibility to evaluate academic influence of academic monographs, so they can only be used to evaluate the social influence of academic monographs. 2020 Vol. 39 (9): 896-905 [Abstract] ( 326 ) HTML (122 KB)  PDF (3116 KB)  ( 664 )

906 Identification of Potential Research Partners Based on Two-Mode Network Analysis Hot! Huang Lu, Ni Xingxing, Cheng Kefei, Jia Xiang DOI: 10.3772/j.issn.1000-0135.2020.09.003 With the increasing complexity and interdisciplinary nature of scientific research, several high-quality research results are found to be obtained owing to the collaboration between researchers. Based on the Web of Science databases, a novel method for identifying and evaluating potential scientific research partners via two-mode network link prediction is developed. The structure of the research content and the cooperative network of the research articles are comprehensively considered. Further, the dynamic nature of research interests and directions are reflected to help researchers rapidly identify potential partners using existing literature. As a part of the empirical research, this study consider scholars in the field of library and information science as an example to recommend partners. 2020 Vol. 39 (9): 906-913 [Abstract] ( 268 ) HTML (160 KB)  PDF (1453 KB)  ( 896 )

914 Ontology Construction for Fire Emergency Management Hot! Wang Fang, Yang Jing, Xu Lulu DOI: 10.3772/j.issn.1000-0135.2020.09.004 Fire outbreak is one of the main emergencies that threaten lives and properties. The number of reports of fire emergencies has increased exponentially in the big data era. Determining how to acquire, extract, and represent knowledge from a large set of fire reports and forming a fire emergency knowledge base to guide the intelligent development of fire emergency management work have become key factors in the effort to improve China's emergency management capability. This paper constructed an ontology model called FEO for fire emergency management. First, based on the guidance of domain experts, the fire emergency upper body FE-SUMO was constructed; then, the five-tuple of FEO consisting of concepts, relationships, functions, axioms and examples was established from fire emergencies, combustion factors, combustion results, fire emergency organization, fire emergency resources, fire emergency roles, and so on. The FEO was implemented using the ontology construction tool Protégé. Finally, the validity and integrity of the constructed fire emergency ontology was verified by qualitative evaluation and OntoQA quantitative evaluation. 2020 Vol. 39 (9): 914-925 [Abstract] ( 294 ) HTML (112 KB)  PDF (2710 KB)  ( 827 )

926 Factors Influencing Relevance Judgment in Video Retrieval: An Empirical Study Based on PLS Path Analysis Hot! Wang Zhihong, Cao Shujin DOI: 10.3772/j.issn.1000-0135.2020.09.005 Relevance evaluation is a key aspect of video information seeking retrieval. Thus, to design and develop a better support system, it is crucial to investigate the factors influencing humans' relevance judgment in video retrieval . Based on existing theories, this paper constructed a multiple-factor model of relevance judgment in video retrieval. Subsequently, structured data were collected from 56 subjects who completed three simulated tasks and answered pre-task and post-task questionnaires. PLS path analysis was used to analyze the collected data and test the model. The results showed that the measurement model had confirmed reliability and validity. The structural model test revealed that topicality, scope, and authority, respectively, were significant factors, whereas understandability, availability, and video characteristics were found to have no statistically significant impact on relevance judgment in video retrieval. Scope also had a significant influence on topicality, thus influencing relevance judgment indirectly. Group analysis indicated that moderating factors including gender, information search ability, and topic familiarity adjusted the impact of the other influencing factors on relevance judgment. This study demonstrates that the factors influencing relevance judgment have some stability and commonality across different contexts and types of information. These results suggest that video system design and development should consider factors beyond topicality to help users effectively and efficiently access video. 2020 Vol. 39 (9): 926-937 [Abstract] ( 249 ) HTML (156 KB)  PDF (1620 KB)  ( 979 )

938 Information Extraction and Integration of Large-scale Heterogeneous Socio-economic Statistical Statements Hot! Zhao Hong, Wang Fang DOI: 10.3772/j.issn.1000-0135.2020.09.006 To better serve the government and the public, full mining of government statistics such as the National Strategic Gold Mine has become an inevitable requirement for the development of big data systems in current smart e-government and new think tanks. However, it is impossible to directly correlate and aggregate statistics due to the semi-structured and large-scale heterogeneous characteristics of statistical statements, which causes significant difficulties in terms of standardized management, deep mining, and extensive utilization of statistical resources. In view of the deficiencies in existing research, this study defines the processing tasks based on the analysis of the semantic elements of government statistical statements and the application objectives of information extraction and integration. The processing tasks are divided into five logical processes: table semantic structure analysis, header semantic relationship recognition, numerical information extraction and representation, index terminology redundancy conversion, and inconsistent statistical data disambiguation loading and the roles and main tasks of each process are described. Finally, this study investigates and constructs the overall technical framework and processing flow. The processing and application results for large-scale real data sets reveal that this method can effectively solve the research question, and has a certain practical value. At the same time, it can also provide reference for other big data construction and application research based on semi-structured tables. 2020 Vol. 39 (9): 938-948 [Abstract] ( 196 ) HTML (104 KB)  PDF (4270 KB)  ( 753 )

949 Identification Method and Empirical Research of Sleeping Beauty in Science Based on Cumulative Citation Hot! Hou Jianhua, Zhang Xuewen DOI: 10.3772/j.issn.1000-0135.2020.09.007 From the perspective of cumulative citation of literature, using the logistic curve as a prototype to fit the cumulative citation curve model, three stages of the life cycle of literature are given by the characteristics of the accumulated citation curve: the sleeping period, the waking period, and the aging period. From the two dimensions of sleep intensity and recovery intensity, we constructed a cumulative citation sleeping beauty index (Cc index), a non-parametric index method for identifying sleeping beauty literature. The sleeping beauty literature in the field of string theory published between 2008-2014 on the Web of Science (WoS) platform was tested. The results revealed that Cc index recognition results in more effectively satisfying the definition of the sleeping beauty document, and has a better recognition effect for the literature of light sleep (doze) and the literature of sleeping beauty with a relatively short sleep time but strong awakening intensity, and simultaneously, can better identify the all-elements-sleeping-beauties literatures, which is more in line with essential characteristics of the sleeping beauty literature. However, further verification is needed to determine if the cumulative citation sleep beauty index identification method is effective for other research fields. 2020 Vol. 39 (9): 949-962 [Abstract] ( 221 ) HTML (153 KB)  PDF (3030 KB)  ( 636 )

963 Discontinuous Usage Behavior Model of Social Media Users: An Empirical Study Hot! Cheng Huiping, Su Chao, Wang Jianya DOI: 10.3772/j.issn.1000-0135.2020.09.008 Clarifying the factors that determine social media users' discontinuous usage behavior has great significance for social media operators, because it facilitates the optimization of the service structure and reduces the negative impact of social media usage. Using the cognition-affection-conation (CAC) categorization of the cognitive emotion theory (CET) as its model framework, this paper constructs the discontinuous usage behavior model of social media users by integrating expectation disconfirmation theory (EDT) and the negative emotions of social comparison theory (envy, regret, and frustration). The data for this study were obtained from 541 valid responses to a network questionnaire, and the partial least squares structural equation modeling (PLS-SEM) was applied in the empirical analysis. The results are as follows. Expectation disconfirmation will lead to dissatisfaction after users utilize social media. Dissatisfaction is the main factor that influences users to form discontinuous usage intention, which will further affect social media users' discontinuous usage behavior. The envy generated by social comparison among users on social media can lead to negative emotions such as frustration and regret. Envy, regret, and frustration will all generate dissatisfaction among social media users. Regret had a significant direct effect on discontinuous usage intention, but the direct effects of frustration are not significant. Expectation disconfirmation can cause users to regret using social media. Regret partially mediates the effects of expectation disconfirmation as well as envy on dissatisfaction. Frustration plays a partially mediating role between envy and dissatisfaction. Dissatisfaction plays a partially mediating role between regret and discontinuous usage intention, and fully mediates the effect of frustration and envy on discontinuous usage intention. The length of social media usage and the number of friends on social media have no significant influence on social media users' discontinuous usage behavior, whereas the length of time spent on social media daily has a significant influence. This study enriches and expands the theoretical research on the discontinuous usage of social media and serves as a practical reference for social media operators to comprehensively understand users' discontinuous usage behavior. 2020 Vol. 39 (9): 963-978 [Abstract] ( 305 ) HTML (197 KB)  PDF (1880 KB)  ( 887 )

979 Identifying and Visualizing Emerging Trends in Domain Based on PWLR Model Hot! Liu Ziqiang, Hu Zhengyin, Xu Haiyun, Fang Shu DOI: 10.3772/j.issn.1000-0135.2020.09.009 Exploring and constructing an accurate and effective framework for the analysis of emerging trends is of great significance to the information work of emerging trends judgment and research, public opinion monitoring, and so on. First, Bi-Gram and Tri-Gram, which are multi-lexical features, are extracted based on an N-Gram model, and a piecewise linear regression (PWLR) model then is used to fit these features. In addition, the model is used to detect the emerging multi-lexical features on the recent time line as well as to accurately identify new words with potential for development and identify new words based on the previous step. A hierarchical clustering algorithm is used to identify emerging trends in the field and visualize the results. Through empirical research, the main emerging trends in the field of gene editing are identified as CRISPR-Cas9 technology, gene therapy, and animal and plant gene editing, which verifies the feasibility and validity of the method proposed herein. 2020 Vol. 39 (9): 979-988 [Abstract] ( 157 ) HTML (129 KB)  PDF (3541 KB)  ( 820 )

989 A Comparative Study between Chinese and Foreign Technology Q&A Communities: Using CSDN and Stack Overflow as Examples Hot! Li Shengli, Zhong Ying DOI: 10.3772/j.issn.1000-0135.2020.09.010 Social question-and-answer (Q&A) communities are important platforms where users share knowledge. Technology Q&A communities focus on technology-related knowledge sharing. This study conducts a comparative research between two large technology Q&A communities at home and abroad to reveal the advantages and shortcomings of the domestic technical Q&A community and provide suggestions for its further development. The comparison focuses on five aspects: community activeness, community topics, community question and answer characteristics, users' experience, and social attributes. In this study, we find the similarities and differences between the two technical Q&A communities. We then provide suggestions for further development and improvement of the Chinese online Q&A community in different aspects. 2020 Vol. 39 (9): 989-1000 [Abstract] ( 213 ) HTML (107 KB)  PDF (3467 KB)  ( 617 )

1001 Comparative Study of Discipline Evaluation Based on Bibliometrics and Topic Detection Considering the Education Disciplines of China, the U.S., the U.K., and Australia Hot! Wang Nan, Ma Qianchun DOI: 10.3772/j.issn.1000-0135.2020.09.011 Since the introduction of the “double first-class”, the academic community have been following the concept that the foundation obtained from a first-class university inculcates discipline. Further, research on discipline evaluation, particularly the evaluation methods and tools employed, has attracted considerable attention. This study compares the productivity and impact of research outcomes and topics in the discipline of education of China (excluding Hong Kong, Macau, and Taiwan), the U.S., the U.K., and Australia from 2013 to 2018, using bibliometrics indicators and research topic detection analyses. The research findings are as follows. A gap still exists between China and the U.S., the U.K, and Australia regarding research competitiveness in the discipline of education. First, compared with the U.S., the U.K., and Australia, the productivity and impact of research outcomes are dissatisfactory in China. Second, the research field is narrow; however, China monitors the global hot spots. Third, China has published the highest number of papers in the topics of engineering talent training and migrant and left-behind children in China. The paper suggests that the bibliometrics indicators of scientific research output and research topic detection methods could be further combined based on the characteristics of different disciplines. In this manner, we can better analyze the advantages and disadvantages of disciplines, evaluate the competitiveness of disciplines, and then serve help develop these disciplines. 2020 Vol. 39 (9): 1001-1010 [Abstract] ( 143 ) HTML (120 KB)  PDF (1234 KB)  ( 595 )