Presentation 2019-03-15
Likability Estimation Model Training of Call-center Agents Based on Annotators' Skills
Hosana Kamiyama, Atsushi Ando, Ryo Masumura, Satoshi Kobashikawa, Yushi Aono,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This paper proposes a new technique for estimating the likability of call-center agents. Most techniques of likability estimation collect many annotations per sample since the annotations often include the variability of annotator evaluations. Due to the inability to collect sufficient annotations, which are labeled by experts such as supervisor of call-center, from actual call-centers, the few annotations per call available cannot eliminate the variability. We proposes two likability model training techniques from the few annotations per call based on annotators’ skill. First, our technique provides reasonable labels, which eliminated from the influence of annotators’ skills, for model training from observed labels. We also proposes a new neural network architecture that has a layer considering the variability of observed labels from each annotator. Our proposal achieved 12.0% error reduction compared with baseline techniques.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Likability / Call-center agent / Annotator / Neural Network
Paper # EA2018-132,SIP2018-138,SP2018-94
Date of Issue 2019-03-07 (EA, SIP, SP)

Conference Information
Committee EA / SIP / SP
Conference Date 2019/3/14(2days)
Place (in Japanese) (See Japanese page)
Place (in English) i+Land nagasaki (Nagasaki-shi)
Topics (in Japanese) (See Japanese page)
Topics (in English) Engineering/Electro Acoustics, Signal Processing, Speech, and Related Topics
Chair Suehiro Shimauchi(Kanazawa Inst. of Tech.) / Shogo Muramatsu(Niigata Univ.) / Yoichi Yamashita(Ritsumeikan Univ.)
Vice Chair Kenichi Furuya(Oita Univ.) / Kanji Watanabe(Akita Pref. Univ.) / Naoyuki Aikawa(TUS) / Kazunori Hayashi(Osaka City Univ) / Akinobu Ri(Nagoya Inst. of Tech.)
Secretary Kenichi Furuya(Shizuoka Inst. of Science and Tech.) / Kanji Watanabe(NHK) / Naoyuki Aikawa(Takushoku Univ.) / Kazunori Hayashi(Hiroshima Univ.) / Akinobu Ri(Kyoto Univ.)
Assistant Keisuke Imoto(Ritsumeikan Univ.) / Daisuke Morikawa(Toyama Pref Univ.) / Katsumi Konishi(Hosei Univ.) / hyihsin(Takushoku Univ.) / Tomoki Koriyama(Tokyo Inst. of Tech.) / Satoshi Kobashikawa(NTT)

Paper Information
Registration To Technical Committee on Engineering Acoustics / Technical Committee on Signal Processing / Technical Committee on Speech
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Likability Estimation Model Training of Call-center Agents Based on Annotators' Skills
Sub Title (in English)
Keyword(1) Likability
Keyword(2) Call-center agent
Keyword(3) Annotator
Keyword(4) Neural Network
Keyword(5)
1st Author's Name Hosana Kamiyama
1st Author's Affiliation NTT Corporation(NTT)
2nd Author's Name Atsushi Ando
2nd Author's Affiliation NTT Corporation(NTT)
3rd Author's Name Ryo Masumura
3rd Author's Affiliation NTT Corporation(NTT)
4th Author's Name Satoshi Kobashikawa
4th Author's Affiliation NTT Corporation(NTT)
5th Author's Name Yushi Aono
5th Author's Affiliation NTT Corporation(NTT)
Date 2019-03-15
Paper # EA2018-132,SIP2018-138,SP2018-94
Volume (vol) vol.118
Number (no) EA-495,SIP-496,SP-497
Page pp.pp.197-202(EA), pp.197-202(SIP), pp.197-202(SP),
#Pages 6
Date of Issue 2019-03-07 (EA, SIP, SP)