[Purpose/Significance] Social media has become an essential medium for the spread of online rumors due to its wide audience and rapid dissemination. In light of the harmful effects of online rumors, understanding how to effectively control their spread and timely debunk them holds significant theoretical and practical implications for governing the cyberspace ecology. [Method/Process] This study proposed a method to identify themes in online rumor debunking and to categorize users' emotional attitudes. It determined the number of topics using the Latent Dirichlet Allocation (LDA)algorithm and constructed user-level and community-level echo chamber networks based on the optimal number of topics. The study presented a model for analyzing the echo chamber effect in social media networks that debunk rumors and established the analysis formula and test parameters for this effect. Utilizing data from the"China Eastern Airlines MU5735 crash"online public opinion topic, the study analyzed the identification and emotional distribution of six online rumor debunking themes. It further examined the network structure of user comments and forwarding subnets within the echo chamber network, related to the six debunking themes. The study also calculated the homogeneity of each virtual community theme in the comment and forwarding subnets of the echo chamber effect for each debunking theme. [Result/Conclusion] The study finds that an echo chamber effect exists in the transmission process of debunking rumored information in social networks. The forwarding process exhibits stronger network homogeneity than the comment process, and users who comment show higher ambivalence than those who forward the information.

[Purpose/Significance] The process of knowledge integration and recombination in science and technology often faces a series of tradeoffs choice, but the academic community still lacks understanding of the underlying mechanisms. [Method/Process] This article used a stochastic actor-oriented model for network analysis to analyze the cross-temporal network evolution mechanism of patent cooperation network consisting of 1339 patent inventors from Genentech from 2011 to 2019. [Result/Conclusion] This article discovers the trade-off phenomenon that academic inventors make when publishing patents and papers across boundaries. After academic inventors publish papers, their tendency to collaborate in the patent collaboration network actually decreases. This article analyzes three evolutionary mechanisms by which academic publication affects the structure of patent networks from the perspectives of "small world", "network rootedness", and"network brokerage". This study helps to understand the phenomenon of enterprises engaging in basic research that has gradually attracted attention in recent years, and also provides empirical reference for the current scientific research participation of enterprises.

[Purpose/Significance] Cross-discipline is becoming an important driving force for scientific and technological innovation in our country, and constantly spawning new scientific frontiers. How to identify and measure cross-frontier topics is very important for understanding the subject knowledge flow and grasping the new cross-frontier direction. [Method/Process] From the perspective of combining citation analysis and subject analysis, this paper proposed a method of cross-frontier recognition based on citation-topic dual measure. Firstly, the cross-topic was obtained by using natural language processing, text topic recognition and citation relation matrix. Secondly, the cross-topic was further selected by constructing cross-measurement index based on citation-topic. Finally, multi-dimensional measurement indicators of theme influence, novelty, attraction and popularity were proposed to obtain the cross-frontier. [Result/Conclusion] The results show that this method can effectively identify the hot cross-frontiers, emerging cross-frontiers, potential cross-frontiers and weak cross-frontiers in this field.

[Purpose/Significance] The study analyzes the ecological chain of big data policy in the context of big model era, aiming to improve the effectiveness of big data policy and promote big data sharing. [Method/Process] Based on the new characteristics of big data policy in the big model era, combined with ecological theory, big data life cycle theory and policy life cycle theory, this paper expounded the connotation of big data policy ecological chain, and summarized the new challenges faced by big data policy ecological chain by taking scientific data policy as an example, and deduced the corresponding optimization strategy. [Result/Conclusion] The big data policy ecological chain is a sub-unit of ecology, social ecology and political ecology.It is derived and runs through the whole data life cycle of big data production, dissemination, consumption, decomposition and regeneration under the background of big data, and connects many policy processes such as policy issuance, transmission, acceptance, feedback and reform.It takes the allocation of big data resources and supervision of big data ecological activities as its work content.With the goal of controlling data quality, ensuring data security and promoting data sharing, the double-helix chain ecological structure has interaction, circularity, value, authority, legitimacy, exclusivity and scientificity, but it is facing new challenges such as weak nodes, short chains and sparse networks.

[Purpose/Significance] With the popularization of mobile social media, users' excessive dependence on mobile devices and mobile social media applications has negatively affected social life and personal health, and the issue of digital addiction has become increasingly prominent.Therefore, by elucidating the role relationship and the degree of influence among the factors influencing the digital detoxification behavior of mobile social media users, it helps users to cultivate good digital detoxification behavioral habits and ensures a healthy balance between their normal life and the time they spend on using their digital devices, and at the same time, it also provides theoretical and applied guidance for the study of digital detoxification behavior. [Method/Process] The study used a combination of meta-ethnography and rough set methods to analyze and identify 15 influencing factors.And through the Grey-DEMATEL (Decision Making Trial and Evaluation Laboratory)method, the study elucidated the role relationships and degrees of influence of each index on digital detoxification. [Result/Conclusion] The study finds that perceived cost, perceived harm, outcome expectancy, pain avoidance, self-control, positive thinking ability, time perception, and social influence are the key influencing factors.And based on the results of the analysis, the study put forward suggestions to cultivate digital detoxification behavioral habits of mobile social media users, which would also provide a reference for the management decision of the platform providers and related communities.

[Purpose/Significance] With the development and popularization of information technology, a series of irrational group behavior such as cyberbullying, rumor spreading, and fanaticism has emerged on the Internet, seriously disrupting the orderly process in cyberspace and even the entire society.A systematic and in-depth exploration of online irrational group behavior at the theoretical level is beneficial for understanding and governance. [Method/Process] Firstly, the concept of online irrational group behavior was analyzed according to the view of Socio-Technical Interaction Network.Then, existing relevant studies were reviewed.Finally, the theoretical basis, research perspectives and methods, driving factors and corresponding mechanisms of online irrational group behavior were summarized and extracted, constructing a theoretical framework. [Result/Conclusion] The concept of online irrational group behavior includes three components: group irrationality, human-technology interaction, and social interaction.This study integrates existing research results and constructs a framework, which could promote future research to focus on deeper issues in more diverse ways, and help to conduct governance practices more completely and specifically.

[Purpose/Significance] Software plays an important role in modern scientific research, and efficiently identifying software entities in academic literature is of great significance for deeply recognizing the academic value of software, promoting the sustainable development of software and the balanced development of academic ecosystem. [Method/Process] This study first defined software entities, then constructed a software entity recognition domain corpus based on a program-assisted annotation scheme for small knowledge bases.On the basis of which, this study proposed an improved SciBERT-BiLSTM-CRF-wordMixup model and evaluated the recognition effect of the model. [Result/Conclusion] The experimental results show that the improved model SciBERT-BiLSTM-CRF-wordMixup proposed in this study performs best in the software entity recognition task, with an overall F1 value of 87.5%, indicating that the model is able to efficiently recognize software and its related information entities from the text of academic papers.

[Purpose/Significance] Technology evolution and forecasting is an important tool for science and technology management departments and researchers to reveal complex technological innovation processes and identify potential development opportunities, in the context of profound changes in the international science and technology competition environment and the increasingly urgent need for self-sustainable development of science and technology.By introducing Extenics Theory, a methodology that seeking to solve problems by focusing on the nature of contradictory issues and transformation mechanisms, into technology evolution and prediction analysis, it could strengthen the scientific of technology evolution and forecasting analysis, improve the reliability and validity, and also provide the better insight of technology evolution mechanisms. [Method/Process] In this paper, the study reviewed main methods of technology evolution and forecasting of subject field, typical application of Extenics in knowledge management and innovation.It was qualitatively verified by taking the extension analysis of crop yield estimation in agricultural remote sensing as an example. [Result/Conclusion] The study discusses the significance and potential of Extenics theory for mining the frontier of technology, from theoretical foundation, expression and analysis perspectives, then points out the issues that should be attention in process.And the case study shows the applicability of extension theory in technology evolution analysis and forecasting.

[Purpose/Significance] This paper tries to construct a knowledge graph of technology talents, achieve intelligent analysis and identification of high potential technology talents, and provide decision-making support for technology talent evaluation, talent gradient training, etc. [Method/Process] Starting from the characteristics of young scientific and technological talents, this article integrated multi-source heterogeneous scientific and technological big data to design and form a technological architecture for the knowledge graph of scientific and technological talents.A knowledge graph technology architecture for young scientific and technological talents based on three-layer scientific and technological data governance were constructed from dimensions such as growth experience, scientific research environment, innovation ability, and technological field, forming a data resource pool and portrait system for young talents in Hunan. [Result/Conclusion] This study constructs the knowledge map of young scientific and technological talents in Hunan province, which covers more than 20000 scientific and technological talents entities and 400000 relational data of Hunan Scientific and technological talents management system, and effectively supports the decision-making of young scientific and technological talents management in Hunan province.

[Purpose/Significance] Online surrogate health information seeking has become an important way for information vulnerable groups to bridge the digital divide.From the perspective of substituted searchers, studying the impact of online surrogate health information seeking on health literacy of the elderly has important reference significance for further understanding the behavioral effects of surrogate seeking. [Method/Process] Based on social cognitive theory, this paper constructed the influence model of online surrogate health information seeking on health literacy of the elderly, and put forward five hypotheses.A questionnaire survey was used to collect 238 valid sample data of the elderly.SPSS and SmartPLS software were used for subsequent data analysis and hypothesis test. [Result/Conclusion] Social interaction, online surrogate health information seeking and health consciousness all positively affect the health literacy of the elderly, and the elderly's health consciousness is the most important factor affecting their health literacy.Through the moderating effect analysis, it finds that health consciousness could positively moderate the impact of online surrogate health information seeking on health literacy of the elderly, and social interaction negatively moderate the impact of online surrogate health information seeking on health literacy of the elderly.

[Purpose/Significance] Users' compliance willingness is of great significance for the effectiveness of online health information services and treatment effects.Understanding the key factors and configurational paths of users' compliance willingness is helpful to provide reference for users' self-management practices and the construction and management of online health communities (OHCs). [Method/Process] Based on Elaboration Likelihood Model (ELM)and Technology Acceptance Model (TAM), the model was constructed from dimensions of online health information's characteristics and users' characteristics.This paper conducted a scenario-based experiment.Then, the partial least square-structural equation model (PLS-SEM)was used to analyze the influencing mechanism of various factors on user compliance intention, and the fuzzy set qualitative comparative analysis (fsQCA)was employed to explore the generation path of users' compliance willingness. [Result/Conclusion] Results indicate that argument quality and source credibility have a significantly and positively impact on information usefulness and information ease of use, and positively affect user compliance willingness through the mediating effect of information usefulness and information ease of use.Meanwhile, demander participation and provider competition played a moderating role, and there was a contingency effect of the information type in the process.Moreover, 3 and 5 configurational paths could be utilized to increase high users' compliance willingness of emerging treatment protocols and mature treatment protocols, respectively.The combined approach of PLS-SEM and fsQCA offers a novel methodology for investigating the underlying mechanism that influences the users' compliance willingness of online health information.Moreover, it provides a focused direction and valuable insights for enhancing user compliance.

[Purpose/Significance] The study aims to solve the problem of difficult alignment between aspect entities and comment entities in online health community texts, an end-to-end generative aspect based on sentiment analysis model BERT-WWM-GPT is proposed. [Method/Process] Firstly, in the model training phase, feature vectors containing rich semantic information were extracted from the text using an encoder.Next, based on feature vectors and standard prediction sequences, sentiment triplets were iteratively generated in the decoder, and model parameters were trained through maximum likelihood estimation.Then, in the model inference stage, a prediction sequence was generated in the decoder based on the text semantic feature vector.Finally, used rules to obtain effective sentiment triplet expression. [Result/Conclusion] The results show that the F1 value of BERT-WWM-GPT model is 12.25% and 7.22% higher than that of GTS and MuG RoBERTa-large models in two aspect level sentiment analysis tasks, respectively.BERT-WWM-GPT model can effectively extract multiple affective triples from online health community reviews, and has excellent generalization ability in other fields.

[Purpose/Significance] Ancient literature books, as one of the important categories of ancient book resources, have unique research value.Using associated data technology to study their associated organization and associated data release is helpful to promote their digital research, development and utilization. [Method/Process] On the basis of the existing"ontology model+associated data realization platform"publishing ideas, the paper analyzed the structural elements of ancient literature books resources, proposed the ancient literature books resources associated data release model which included five levels: data layer, data network layer, data fusion layer, application layer and presentation layer, and took the representative ancient literature book Liaozhai Zhiyi · Siwenlang as an example. [Result/Conclusion] The ancient literature books resources associated data release model constructed in this paper integrates GPT technology in the data layer to complete the data collection work, improve the efficiency of knowledge organization, and adds user demand analysis in the application layer to improve the practicality of the model and the effect of associated data publishing.The experimental results show that the publishing model constructed in this paper can meet the needs of related publishing of ancient literature books resources, can deeply reveal the semantic connotation of such ancient literature books, and provide a feasible implementation plan for ancient literature books from data collection to publication.At the same time, it can provide related knowledge services of ancient literature books resources for users' needs.

[Purpose/Significance] Oral historical resources have important historical and spiritual value, with heterogeneous resource characteristics across multiple fields such as libraries, archives, museums, and science and technology museums.The historical events of celebrities separated from oral historical resources, as well as their new"bottom-up"approach to studying history, have received widespread attention in the fields of history, archives, and information.The organization and application of event knowledge is the key and difficult issue in knowledge processing, and from the perspective of digital humanities, the knowledge-based transformation of celebrity historical events is an important foundation for knowledge reorganization, value mining, and narrative performance of celebrity special collection resources. [Method/Process] On the basis of previous research, the article summarized and analyzed the connotation, characteristics, and application requirements of celebrity historical events, constructed a semantic model of celebrity historical events, and instantiated the semantic model using the scientist Li Zhengdao and the founder CUSPEA event as an example. [Result/Conclusion] Based on the semantic model, the graph database could form more flexible, fine-grained, scalable, and related entity relationships and knowledge, achieving the application requirements of semantic level queries, thematic aggregation, narrative display, and visual presentation of celebrity characteristic resources for different types of users.