Hidden Communities in the Russian Social Network Corpus: a Comparative Study of Detection Methods

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

The paper presents a comparative study of different methods that
help to detect hidden communities within social networks. The tested approaches were divided into three main groups: a graph-based method, a clustering method, and a hybrid method. The experiments were conducted on the Russian corpus of posts from VKontakte social network. We discuss advantages and
disadvantages of all the methods, and predict the ways of their improving.
Original languageEnglish
Title of host publicationProceedings of the Computational Models in Language and Speech Workshop (CMLS 2020) co-located with 16th International Conference on Computational and Cognitive Linguistics (TEL 2020), Kazan, Russia, November 12-13, 2020
Pages69-78
StatePublished - 18 Dec 2020
EventComputational Models in Language and Speech Workshop (CMLS 2020)
co-located with 16th International Conference on Computational and Cognitive Linguistics (TEL 2020)
- Kazan, Russian Federation
Duration: 12 Nov 202013 Nov 2020

Publication series

NameCEUR Workshop Proceedings
Volume2780

Conference

ConferenceComputational Models in Language and Speech Workshop (CMLS 2020)
co-located with 16th International Conference on Computational and Cognitive Linguistics (TEL 2020)
Abbreviated titleCMLS 2020, TEL 2020
Country/TerritoryRussian Federation
CityKazan
Period12/11/2013/11/20

Keywords

  • Social Networks
  • Corpus Linguistics
  • Jaccard Index
  • Cluster analysis
  • Topic Modeling
  • Automatic topic labeling

Fingerprint

Dive into the research topics of 'Hidden Communities in the Russian Social Network Corpus: a Comparative Study of Detection Methods'. Together they form a unique fingerprint.

Cite this