Skip to the main content

Preliminary communication

https://doi.org/10.17559/TV-20210610120531

A Hotspot Discovery Method Based on Improved FIHC Clustering Algorithm

Lina Lin ; Information Engineering School, Jimei University Chengyi College, Xiamen 361021, China
Dezhi Wei* orcid id orcid.org/0000-0001-6413-9309 ; Information Engineering School, Jimei University Chengyi College, Xiamen 361021, China


Full text: english pdf 561 Kb

page 1790-1796

downloads: 430

cite


Abstract

It was difficult to find the microblog hotspot because the characteristics of microblog were short, rapid, change and so on. A microblog hotspot detection method based on MFIHC and TOPSIS was proposed in order to solve the problem. Firstly, the calculation of HowNet similarity was used in the score function of FIHC, the semantic links between frequent words were considered, and the initial clusters based on frequent words were produced more accurately. Then the initial cluster of the text repletion of mircoblog was reduced, and the idea of Single-Pass clustering was used to the reduced topic cluster in order to get the Hotspot. At last, an improved TOPSIS model was used to sort the hot topics in order to get the rank of the hot topics. Compared with the other text clustering algorithms and hotspot detection methods, the method has good effect, and can be a more comprehensive response to the current hot topics.

Keywords

clustering; hotspot detection; internet public opinion; microblog; TOPSIS

Hrčak ID:

261360

URI

https://hrcak.srce.hr/261360

Publication date:

15.8.2021.

Visits: 1.188 *