341 The Diffusion of Citations for Individual Publications Min Chao, Ding Ying, Li Jiang, Sun Jianjun DOI: 10.3772/j.issn.1000-0135.2018.04.001 Citation is an important connection that links scientific knowledge, providing us with a key to understanding the acquisition, communication, transformation, and innovation of knowledge. Citations are not only static statistic distributions, but also dynamic growth processes in the complex system of modern scientific communication. Considering citation as a diffusion process, this paper reconsiders the essential elements, quantification metrics, and growth mechanisms of this process from multiple approaches. The citation diffusion system involves at least these elements: time environment, time of first citation, citation peak, total citation counts, citation rate, citation lifecycle, literature network/ecosystem, and citation “takeoff”. As a diffusion unit on the micro scale, citation cascade evolves from the two dimensions of cascade virality and cascade popularity, exhibiting two typical patterns of broadcast diffusion and viral diffusion. The huge diversity of diffusion patterns stems from the underlying mechanisms that govern citation diffusion, which may include preferential attachment, aging, fitness, as well as innovation and imitation mechanisms. Further exploration of citation diffusion and its mechanisms will not only enhance our understanding of this process, but also help reveal its potential value in such scientific activities as research evaluation, scientific prediction, and scientific policy-making. 2018 Vol. 37 (4): 341-350 [Abstract] ( 240 ) HTML (1 KB)  PDF (2621 KB)  ( 879 )

351 Citation Network Main Path Identification Based on Associated Attributes of Articles: Case Study from Synthetic Biology Wei Ling, Liu Chunjiang, Xu Haiyun, Fang Shu DOI: 10.3772/j.issn.1000-0135.2018.04.002 State-of-the-art citation network main path analysis methods either treat all citation relationships with equal weightage, or calculate the relevance of citations based on the content of the text. Neither nonequivalence of citation nodes nor relevance of citations based on associated attributes of articles are considered. This study uses meta paths to describe and quantify the relevance of citations based on associated attributes of articles, considering it as part of the main path traversal weight, and combines it with SPC (search path count) to construct two new indices to analyze the contribution of associated attributes of articles to main path identification. The results prove that the new traversal weight indices can identify main path characteristics of associated attributes, reveal knowledge diffusion paths in different views, and provide detailed diffusion information. This work expands the functions and application scenarios of citation network main path analysis. 2018 Vol. 37 (4): 351-361 [Abstract] ( 332 ) HTML (1 KB)  PDF (822 KB)  ( 838 )

362 An Empirical Study on Privacy Policy Reading Intention of Social Media Users Zhu Hou, Zhang Mingxin, Lu Yonghe DOI: 10.3772/j.issn.1000-0135.2018.04.003 Privacy policies are sets of terms that inform users of how their personal information is collected and utilized by social media platforms. However, in the actual process, users rarely or never read privacy policies. In explaining this phenomenon, suggestions are made for the formulation and improvement of privacy policies and the protection of users’ personal information. Based on the relevant literature at home and abroad, subjective norms, motivation, reading ability, content integrity, and readability are introduced. The structural equation model is used to test the hypothesis on 344 samples. On the basis of the test results, the content integrity and readability of seven widely used social media privacy policies are analyzed. Subjective norms, motivation, readability, content integrity, and readability all have positive effects on users’ reading intentions. Privacy policies characterized by specific content, transparency, and reasonable length will have higher reading rates. 2018 Vol. 37 (4): 362-371 [Abstract] ( 328 ) HTML (1 KB)  PDF (427 KB)  ( 1187 )

372 The Impact of Group Heterogeneity on Knowledge Ordering Efficiency in Online Knowledge Communities Qiu Jiangnan, Wang Jingxian DOI: 10.3772/j.issn.1000-0135.2018.04.004 In the Web 2.0 era, the online knowledge community (OKC) has become an important platform for producing, sharing, and acquiring knowledge. How to realize knowledge ordering and provide high-quality information and knowledge more efficiently is a matter of concern for the OKC. In this paper, a theoretical model of factors affecting OKC knowledge ordering efficiency based on interactive team cognitive (ITC) theory is proposed in order to investigate how group heterogeneity and team interaction influence knowledge ordering efficiency. In this paper, we chose English Wikipedia as the research platform and use a random sample of 100 featured articles acquired through its “random featured article” function. We obtained objective data about the articles and editors through an application program interface (API). The partial least squares (PLS) path modeling was used to validate the theoretical model. Results suggest that knowledge heterogeneity and experience heterogeneity have no significant direct effect on knowledge ordering efficiency, but that knowledge heterogeneity has a positive impact on knowledge collaboration and cognitive conflict whereas experience heterogeneity has a negative impact on cognitive conflict. Knowledge collaboration and cognitive conflict have a positive impact on communication, and there is a non-linear relationship between communication and knowledge ordering efficiency. Our study not only makes a theoretical contribution to OKC knowledge ordering research, but also provides practical implications for the design of collaboration tools and management of OKCs. 2018 Vol. 37 (4): 372-383 [Abstract] ( 168 ) HTML (1 KB)  PDF (490 KB)  ( 603 )

384 Study of a Group Recommendation Model of Integrating Context Information in a Mobile Environment—Empirical Analysis Based on User APP Behavior Data Xia Lixin, Yang Jinqing, Cheng Xiufeng DOI: 10.3772/j.issn.1000-0135.2018.04.005 In order to improve the accuracy of recommended results in the group recommendation model, a group recommendation model integrating context information is proposed in this paper. Firstly, the user behavior context data are obtained, and the preference represented by individual user behavior is extracted. Secondly, the behavior similarity of individual users is calculated and cluster discovery is conducted. Subsequently, the group behavior characteristics are mined from context data, and then a feature vector of group behavioral preference is built. Finally, collaborative recommendation ideas are combined for the group as a whole. The collaboration also occurs with other groups producing an item history score to form a prediction score. In the experiment, we analyze the user’s operation flow, extract the theme sequence features, and then incorporate the classic context information to produce the recommendation results. The results show that the top-6 of the recommended results obtained by using this model are more accurate than those recommended by traditional (non-situational) groups. Therefore, this model is more suitable for group recommendations in mobile environments. 2018 Vol. 37 (4): 384-393 [Abstract] ( 148 ) HTML (1 KB)  PDF (1124 KB)  ( 681 )

394 Concerns and Evolutionary Patterns of Stakeholders on Social Media Platforms during Public Health Emergencies An Lu, Du Tingyao, Li Gang, Yu Chuanming DOI: 10.3772/j.issn.1000-0135.2018.04.006 In this study, a system of stakeholders of public health emergencies is constructed based on stakeholder theory, which consists of thirteen types. The life cycle of emergencies is divided into five phases, based on life cycle theory. The Latent Dirichlet Allocation (LDA) model is used to analyze and compare the similarities and differences of topics about the Middle East Respiratory Syndrome (MERS) outbreak on Sina Weibo and WeChat platforms. The concerns and evolutionary patterns of stakeholders in the different life cycle phases are characterized and summarized. The findings can help public health emergency administrators take corresponding measures for specific populations and alleviate the negative impact of public emergencies. 2018 Vol. 37 (4): 394-405 [Abstract] ( 358 ) HTML (1 KB)  PDF (1991 KB)  ( 2141 )

406 Text Mining on the Government Work Reports of the State Council (1954-2017) and Social Transformation Research Wei Wei, Guo Chonghui, Chen Jingfeng DOI: 10.3772/j.issn.1000-0135.2018.04.007 The government work report of the State Council is a comprehensive policy text. This paper uses text mining technology to carry out a comprehensive multi-granularity, multi-level quantitative analysis of the government work reports. This has immense practical significance for acquiring an understanding of the evolution of domain content and the laws of social transformation discovery to relevant personnel. Firstly, a series of text preprocessing is done by using the Chinese word segmentation tool combined with three kinds of dictionaries created by us: the domain dictionary, the stop word dictionary, and the thesaurus dictionary. Then, frequent words, hot words, and new words are redefined and three kinds of corresponding feature mining methods are proposed. A quantitative calculation method for social vitality is proposed based on these new words, and then clustering analysis is conducted for feature words represented by a time series with a popular clustering method. According to the document information of the government work reports, we divide the time period from 1954 to 2017 into different stages, using which we conduct the pattern discovery for feature words combined with the feature words clustering results. Finally, our findings show that the selected frequent words, hot words, and new words in the government work reports can indicate the common problems, the hot issues and its evolution pattern, and the changes in social activity over the years. From the term clustering results and the reasonable time stages of the whole period, we get nine specific patterns of feature words. 2018 Vol. 37 (4): 406-421 [Abstract] ( 298 ) HTML (1 KB)  PDF (2062 KB)  ( 1103 )

422 A Review of the Cross-Modal Retrieval Model and Feature Extraction Based on Representation Learning Li Zhiyi, Huang Zifeng, Xu Xiaomian DOI: 10.3772/j.issn.1000-0135.2018.04.008 Representation learning, particularly deep learning, has received wide attention and seen application in speech recognition, image analysis, and natural language processing fields. It not only promotes the research and development of artificial intelligence, but urges enterprises to consider new business and profit models. This paper aims to examine these studies in the form of reviews, and ultimately form a complete overview of the topic. Through the investigation and organization of relevant literature locally and internationally, this paper summarizes the research results of cross-modal retrieval and feature extraction based on representation learning from the two dimensions of information extraction and representation, and cross-modal system modeling. The main research includes summarizing five traditional representation learning algorithms, which are the autoencoder, sparse encoding, the restricted Boltzmann machine, deep belief networks, and convolutional neural networks. From the shared layer relationship between each mode, the representation space, and the correlation between each mode’s in-depth learning-based cross-modal modeling algorithm, the present state of research on modeling systems based on cross- modal modeling is summed up. Finally, the evaluation index of cross-modal retrieval is summarized. The study finds that the existing retrieval research is rich in single-modal information retrieval and that the content of queries and candidate sets belong to the same modality, whereas cross-modal retrieval is limited to two modal alignment languages of images and texts. Future research needs to see an increase of modal retrieval of audio, video, images, text, and other multimodal data, and using deeper constructing multimodal retrieval models and feature extraction algorithms to achieve three-or- greater cross-modal retrieval. In addition, an evaluation index of multimodal retrieval systems must be established. 2018 Vol. 37 (4): 422-435 [Abstract] ( 247 ) HTML (1 KB)  PDF (841 KB)  ( 2616 )

436 Review of Clickstream Data Analysis and Visualization Studies Jiang Tingting, Xu Yaping, Guo Qian DOI: 10.3772/j.issn.1000-0135.2018.04.009 This review provides an overview of studies on clickstream data analysis and visualization performed between 2000 and 2017. In-depth analysis and comparisons have been performed on the visual characteristics of visualization studies available in literature. The analysis reveals that the existing visualizations (1) usually involve users’ visiting footprints, movements, and pathways as components of representation, (2) present them as directed graphs, sequence diagrams, line graphs, network graphs, cloud maps, Sankey diagrams, matrix graphs, stack graphs, and multiply graphs, and (3) support user interaction through features such as overview, zoom, filter, details-on-demand, relate, history, and extract. This review suggests that future studies should focus on the application of big data technologies in clickstream data analysis, visualization of mobile clickstream data, and user testing for visualization evaluation. 2018 Vol. 37 (4): 436-450 [Abstract] ( 172 ) HTML (1 KB)  PDF (3485 KB)  ( 1069 )