|
|
|
| On the Theme of Network Science Conference via Text Mining |
| LI Xiaoke1, ZHAO Zijuan1, GUO Qiang1, LIU Jianguo2, LI Rende1
|
1. Complex Systems Science Research Center, University of Shanghai for Science and Technology, Shanghai 200093, China;
2. Fintech Institute, Shanghai University of Finance and Economics, Shanghai 200433, China |
|
|
|
Abstract The rapid development of network science has had a profound impact on the disciplines of physics, computers and management. However, the current development trend of the latest network science topics in China has been lack of intuitive analysis. This paper takes the summary of the 13th National Network Science Conference in 2017 as the research object, and analyzes the research trend of the most representative complex network conferences in network science from the perspective of topic extraction and clustering based on text mining. The research trend can reflect the latest research situation in the field of domestic network science to a certain extent. Firstly, the text information of the conference summary is preprocessed, and the text is jieba word segmentation through the self-built dictionary and the stop vocabulary. Then use the LDA topic model to identify the topic distribution of the abstract, and perform based on the JS distance between the abstracts to get 10 types of conference topics. This paper expands the application scope of the topic model in the research situation and research hotspots of academic conferences, enriches the ideas of academic conference topic mining and research hotspot analysis, and can provide reference for other academic conferences to quickly explore research situation. At the same time, it proposes a kind of combination of topic model and social network analysis, this paper explores the research methods of institutional associations, and uses the similarity of institutional research topics as reference indicators to provide reference for institutions to find suitable research cooperation units.
|
|
Received: 28 June 2018
Published: 31 January 2019
|
|
|
|
|
|
| No related articles found! |
|
|
|
|
