Details

Autor(en) / Beteiligte

• Huang, Yujian
• Liu, Qiang
• Liu, Jia
• Hu, Yanmei

Titel

Topic Discovery in Scientific Literature

Ist Teil von

• Computer Supported Cooperative Work and Social Computing, p.481-491

Ort / Verlag

Singapore: Springer Nature Singapore

Link zum Volltext

Quelle

Alma/SFX Local Collection

Beschreibungen/Notizen

• With the progress of society and science, various fields have achieved unprecedented development. Various research directions and problems have blossomed, and a huge amount of scientific literature has emerged. Scientific literature contains rich “knowledge”, e.g., research hotspots and topics. If those “knowledge” can be obtained from scientific literature, it would be of great practical significance to both government and researchers. The existing methods generally obtain “knowledge” by analyzing the semantics of scientific literature, which is complex and time-consuming. In this paper, we aim to explore new methods of research hot word extraction and research topic discovery from the perspective of network. Firstly, the word network is constructed based on the text of scientific literature. Next, a research hot word extraction method based on node centrality and a structural topic discovery method are proposed on the word network. Then, the consistency between structural topics and semantic topics is explored. Finally, the proposed methods are experimentally verified on a real dataset. The experimental results show that the proposed centrality based hot word extraction method can effectively extract research hot words, and the topics obtained by the structural topic discovery method are consistent with the semantic topics in some cases, providing a new way to textual knowledge discovery.