Presentation 2018-06-28
Analysis of solution diversity about topic model
Toshio Uchiyama,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) A Topic model is known as a useful statistical model for analyzing text data and images. Many different parameters (solutions) of a topic model can be obtained due to different initial conditions. To utilize diversity of solutions, we proposed the method to assign coordinate values to the solutions by defining inner product (similarity) between solutions, where inner product is normalized mutual information between them. Through the experiment for one document data set, usefulness of the proposed analysis method had been shown. In this report, we apply it to various document data and analyze it, and show the experimental results and investigate them.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) topic model / diversity of solutions / normalized mutual information
Paper # PRMU2018-26,SP2018-6
Date of Issue 2018-06-21 (PRMU, SP)

Conference Information
Committee PRMU / SP
Conference Date 2018/6/28(2days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair Shinichi Sato(NII) / Yoichi Yamashita(Ritsumeikan Univ.)
Vice Chair Yoshihisa Ijiri(Omron) / Toru Tamaki(Hiroshima Univ.) / Akinobu Ri(Nagoya Inst. of Tech.)
Secretary Yoshihisa Ijiri(NEC) / Toru Tamaki(Osaka Univ.) / Akinobu Ri(Kyoto Univ.)
Assistant Go Irie(NTT) / Yoshitaka Ushiku(Univ. of Tokyo) / Tomoki Koriyama(Tokyo Inst. of Tech.) / Satoshi Kobashikawa(NTT)

Paper Information
Registration To Technical Committee on Pattern Recognition and Media Understanding / Technical Committee on Speech
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Analysis of solution diversity about topic model
Sub Title (in English)
Keyword(1) topic model
Keyword(2) diversity of solutions
Keyword(3) normalized mutual information
1st Author's Name Toshio Uchiyama
1st Author's Affiliation Hokkaido Information University(HIU)
Date 2018-06-28
Paper # PRMU2018-26,SP2018-6
Volume (vol) vol.118
Number (no) PRMU-111,SP-112
Page pp.pp.25-30(PRMU), pp.25-30(SP),
#Pages 6
Date of Issue 2018-06-21 (PRMU, SP)