This study explores new ways of tag-based personalized recommendation by relieving the assumption that tags assigned by a user occur independently of each other. The new methods profile users using tag co-occurrence networks, upon which link-based node weighting methods (e.g. PageRank and HITS) are applied to refine the weights of tags. A diffusion process is then performed upon the tag-item bipartite graph to transform the weights of tags into recommendation scores for items. Experiments on three datasets showed improvements of the proposed method over the tag-based collaborative filtering in terms of precision and recall in the datasets with dense user-tag networks and in terms of inter-diversity in all datasets. In addition, the user-tag network is found to be a necessary instrument for the improvements. The findings of this study contribute to more accurate user profiling and personalized recommendations using network theory and have practical implications for tag-based recommendation systems.

Purpose: The purpose of this paper is to identify salient topic categories and outline their evolution patterns and temporal trends in microblogs on a public health emergency across different stages. Comparisons were also examined to reveal the similarities and differences between those patterns and trends on microblog platforms of different languages and from different nations. Design/methodology/approach: A total of 459,266 microblog entries about the Ebola outbreak in West Africa in 2014 on Twitter and Weibo were collected for nine months after the inception of the outbreak. Topics were detected by the latent Dirichlet allocation model and classified into several categories. The daily tweets were analyzed with the self-organizing map technique and labeled with the most salient topics. The investigated time span was divided into three stages, and the most salient topic categories were identified for each stage. Findings: In total, 14 salient topic categories were identified in microblogs about the Ebola outbreak and were summarized as increasing, decreasing, fluctuating or ephemeral types. The topical evolution patterns of microblogs and temporal trends for topic categories vary on different microblog platforms. Twitter users were keen on the dynamics of the Ebola outbreak, such as status description, secondary events and so forth, while Weibo users focused on background knowledge of Ebola and precautions. Originality/value: This study revealed evolution patterns and temporal trends of microblog topics on a public health emergency. The findings can help administrators of public health emergencies and microblog communities work together to better satisfy information needs and physical demands by the public when public health emergencies are in progress.

Purpose: Interdisciplinary fields have become the driving force of modern science and a significant source of scientific innovation. However, there is still a paucity of analysis about the essential characteristics of disciplines’ cross-disciplinary impact. Design/methodology/approach: In this study, we define cross-disciplinary impact on one discipline as its impact to other disciplines, and refer to a three-dimensional framework of variety-balance-disparity to characterize the structure of cross-disciplinary impact. The variety of cross-disciplinary impact of the discipline was defined as the proportion of the high cross-disciplinary impact publications, and the balance and disparity of cross-disciplinary impact were measured as well. To demonstrate the cross-disciplinary impact of the disciplines in science, we chose Microsoft Academic Graph (MAG) as the data source, and investigated the relationship between disciplines’ cross-disciplinary impact and their positions in the Hierarchy of Science (HOS). Findings: Analytical results show that there is a significant correlation between the ranking of cross-disciplinary impact and the HOS structure, and that the discipline exerts a greater cross-disciplinary impact on its neighboring disciplines. Several bibliometric features that measure the hardness of a discipline, including the number of references, the number of cited disciplines, the citation distribution, and the Price index have a significant positive effect on the variety of cross-disciplinary impact. The number of references, the number of cited disciplines, and the citation distribution have significant positive and negative effects on balance and disparity, respectively. It is concluded that the less hard the discipline, the greater the cross-disciplinary impact, the higher balance and the lower disparity of cross-disciplinary impact. Research limitations: In the empirical analysis of HOS, we only included five broad disciplines. This study also has some biases caused by the data source and applied regression models. Practical implications: This study contributes to the formulation of discipline-specific policies and promotes the growth of interdisciplinary research, as well as offering fresh insights for predicting the cross-disciplinary impact of disciplines. Originality/value: This study provides a new perspective to properly understand the mechanisms of cross-disciplinary impact and disciplinary integration.

Along with the rapid development of new network technologies such as cloud computing, the server sides were becoming more and more untrustworthy, and the problem of user privacy has become one of the main obstacles to the further development of modern information services. In this paper, under an architecture of edge computing, we propose to replace the identity associated with each user information service request with an identity from other users on an edge-side, to break the association between users and their requests, and thus protect user information service privacy. First, we present a basic framework under edge computing and identity replacement for the protection of user information service privacy. It requires no changes to the accuracy, efficiency and generality of information services, and no changes to the information service algorithm and user usage habits, so it has good practicability. Second, we define an identity replacement-based model for the protection of user information service privacy, which formulates the privacy constraints and cost constraints that should be satisfied in the process of user identity replacement on an edge-side. Third, according to four matching strategies proposed based on the privacy model, we present an implementation algorithm for the privacy model. Finally, both theoretical analysis and experimental evaluation demonstrate the effectiveness of the proposed approach, i.e., compared with existing technical methods, it has better overall performance in terms of accuracy, efficiency, generality and practicability, which can improve the security of user information service privacy on the untrusted cloud-side, without compromising the usability of information services.