Presentation 2020-06-27
Estimation of Sports Broadcast Situations based on Character-level Auto Encoder for Live Tweets
Nodoka Fujimoto, Taketoshi Ushiama,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this paper, we propose a method to generate vectors of each situation on video content from the latest small number of live tweets in a professional baseball broadcast. In order to vectorize the live tweets appropriately, we use a character-level auto-encoder that can deal with text noise and unique expressions of tweets. Then, to evaluate the usefulness of the proposed method, this paper present the experimental results based on the task of estimating the number of tweet posts that are likely to be correlated with the level of excitement of audiences.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) SNS / Live Tweet / Auto Encoder / Deep Learning
Paper # DE2020-10
Date of Issue 2020-06-20 (DE)

Conference Information
Committee DE
Conference Date 2020/6/27(1days)
Place (in Japanese) (See Japanese page)
Place (in English) Online
Topics (in Japanese) (See Japanese page)
Topics (in English) Social Computing
Chair Jun Miyazaki(Tokyo Inst. of Tech.)
Vice Chair Shohei Yokoyama(Tokyo Metropolitan Univ.) / Kazuo Goda(Univ. of Tokyo)
Secretary Shohei Yokoyama(NTT) / Kazuo Goda(Univ. of Hyogo)
Assistant Saneyasu Yamaguchi(Kogakuin Univ.) / Shoko Wakamiya(NAIST)

Paper Information
Registration To Technical Committee on Data Engineering
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Estimation of Sports Broadcast Situations based on Character-level Auto Encoder for Live Tweets
Sub Title (in English)
Keyword(1) SNS
Keyword(2) Live Tweet
Keyword(3) Auto Encoder
Keyword(4) Deep Learning
1st Author's Name Nodoka Fujimoto
1st Author's Affiliation Kyushu University(Kyushu Univ.)
2nd Author's Name Taketoshi Ushiama
2nd Author's Affiliation Kyushu University(Kyushu Univ.)
Date 2020-06-27
Paper # DE2020-10
Volume (vol) vol.120
Number (no) DE-78
Page pp.pp.53-58(DE),
#Pages 6
Date of Issue 2020-06-20 (DE)