Presentation 2016-09-16
A study on the Multiple Label Propagation Algorithm Considering Properties of the LOD
Toshitaka Maki, Kazuki Takahashi, Toshihiko Wakahara, Akihiro Yamaguchi, Akihisa Kodate, Toru Kobayashi, Sonehara Noboru,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) The label propagation is a semi-supervised learning method to grant to classification labels to the nodes under the assumption of belonging to the same class. The authors noted that there is a problem in the definition of the link relationships and properties of the LOD in the CityData site. We proposed a multiple label propagation algorithm which extends the label propagation in order to solve the problem. The proposed algorithm can realize the extension of the LOD based on the teacher's data. In the present study, the multiple label propagation algorithm has been taken into account the properties of the LOD in order to realize the effective expansion of links and nodes. In addition, this multiple label propagation algorithm is applied to the link prediction of LOD and the link distance has been approached by the updated link. Results of evaluation experiments using I-Scover data show that the label correct answer rate in the case of using a multiple propagation algorithm is improved and the validity of the proposed algorithm has been confirmed.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Linked Open Data / Multiple Label Propagation Algorithm / Graph Prediction / Property / I-Scover
Paper # LOIS2016-24,IE2016-61,EMM2016-50
Date of Issue 2016-09-08 (LOIS, IE, EMM)

Conference Information
Committee ITE-ME / IE / EMM / LOIS / IEE-CMN
Conference Date 2016/9/15(2days)
Place (in Japanese) (See Japanese page)
Place (in English) Aichi Prefectural University
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair Miki Haseyama(Hokkaido Univ.) / Seishi Takamura(NTT) / Akinori Ito(Tohoku Univ.) / Hiroyuki Nishi(Sojo Univ.) / Yoshimi Fukumura(Nagaoka Univ. of Tech.)
Vice Chair Norio Tagawa(Tokyo Metropolitan Univ.) / Takayuki Hamamoto(Tokyo Univ. of Science) / Atsuro Ichigaya(NHK) / Masaki Kawamura(Yamaguchi Univ.) / Hirohisa Hioki(Kyoto Univ.) / Tomohiro Yamada(NTT)
Secretary Norio Tagawa(NEC) / Takayuki Hamamoto(NTT) / Atsuro Ichigaya(Chiba Inst. of Tech.) / Masaki Kawamura(Nagasaki Univ.) / Hirohisa Hioki(Osaka Pref. Univ.) / Tomohiro Yamada(NTT) / (Nagasaki Univ.)
Assistant / Kei Kawamura(KDDI R&D Labs.) / Keita Takahashi(Nagoya Univ.) / Rui Shogenji(Shizuoka Univ.) / Masaaki Fujiyoshi(Tokyo Metropolitan Univ.) / Yukihiro Nakamura(NTT) / Keisuke Utsu(Tokai Univ.) / Takeshi YAGI(NTT)

Paper Information
Registration To Technical Group on Media Engineering / Technical Committee on Image Engineering / Technical Committee on Enriched MultiMedia / Technical Committee on Life Intelligence and Office Information Systems / Technical Group on Communications
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A study on the Multiple Label Propagation Algorithm Considering Properties of the LOD
Sub Title (in English)
Keyword(1) Linked Open Data
Keyword(2) Multiple Label Propagation Algorithm
Keyword(3) Graph Prediction
Keyword(4) Property
Keyword(5) I-Scover
1st Author's Name Toshitaka Maki
1st Author's Affiliation Fukuoka Institute of Technology(FIT)
2nd Author's Name Kazuki Takahashi
2nd Author's Affiliation Fukuoka Institute of Technology(FIT)
3rd Author's Name Toshihiko Wakahara
3rd Author's Affiliation Fukuoka Institute of Technology(FIT)
4th Author's Name Akihiro Yamaguchi
4th Author's Affiliation Fukuoka Institute of Technology(FIT)
5th Author's Name Akihisa Kodate
5th Author's Affiliation Tsuda College(Tsuda)
6th Author's Name Toru Kobayashi
6th Author's Affiliation Nagasaki University(Nagasaki Univ)
7th Author's Name Sonehara Noboru
7th Author's Affiliation National Institute of Informatics(NII)
Date 2016-09-16
Paper # LOIS2016-24,IE2016-61,EMM2016-50
Volume (vol) vol.116
Number (no) LOIS-220,IE-221,EMM-222
Page pp.pp.49-54(LOIS), pp.49-54(IE), pp.49-54(EMM),
#Pages 6
Date of Issue 2016-09-08 (LOIS, IE, EMM)