Skip to main navigation Skip to search Skip to main content

A novel topic number selecting algorithm for topic model

  • Linlin Tang*
  • , Liang Zhao
  • *Corresponding author for this work

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

Abstract

A novel algorithm named the MTN (Multiple-Topic-Number) algorithm is introduced to deal with the problem of topic number selecting in topic model issue. The purpose of our algorithm is to build the LDA (Latent Dirichlet Allocation) matrices of different topic numbers to make the LDA matrices and machine learning algorithm combined better. So it can be used to solve the traditional problem of selecting topic number: under-size or over-size. The method here is to use different levels of machine learning tree structure to complete the combination. Experimental results show the efficiency of our proposed algorithm.

Original languageEnglish
Title of host publicationGenetic and Evolutionary Computing - Proceedings of the 13th International Conference on Genetic and Evolutionary Computing, 2019
EditorsJeng-Shyang Pan, Yongquan Liang, Jerry Chun-Wei Lin, Shu-Chuan Chu
PublisherSpringer
Pages483-490
Number of pages8
ISBN (Print)9789811533075
DOIs
StatePublished - 2020
Externally publishedYes
Event13th International Conference on Genetic and Evolutionary Computing, ICGEC 2019 - Qingdao, China
Duration: 1 Nov 20193 Nov 2019

Publication series

NameAdvances in Intelligent Systems and Computing
Volume1107 AISC
ISSN (Print)2194-5357
ISSN (Electronic)2194-5365

Conference

Conference13th International Conference on Genetic and Evolutionary Computing, ICGEC 2019
Country/TerritoryChina
CityQingdao
Period1/11/193/11/19

Keywords

  • LDA
  • Topic model
  • Xgboost

Fingerprint

Dive into the research topics of 'A novel topic number selecting algorithm for topic model'. Together they form a unique fingerprint.

Cite this