CCSS School on Computational Social Science『計算社会科学入門』における第5章ネットワーク解説.
http://ccss.kobe-u.ac.jp/event/seminar_all/2020/202102270900.html
「計算社会科学」
https://amzn.to/3pZGt2w
A brief presentation on the session "AI Network and Ethics" of the International Forum Toward AI Network Society held by Japan’s Ministry of Internal Affairs and Communications.
CCSS School on Computational Social Science『計算社会科学入門』における第5章ネットワーク解説.
http://ccss.kobe-u.ac.jp/event/seminar_all/2020/202102270900.html
「計算社会科学」
https://amzn.to/3pZGt2w
A brief presentation on the session "AI Network and Ethics" of the International Forum Toward AI Network Society held by Japan’s Ministry of Internal Affairs and Communications.
2. 発表論文
[Title]
Graph Neural Networks for Social Recommendation
[Author]
Wenqi Fan, Yua Ma, Qing Li, Yuan He, Eric Zhao, Jiling Tang, and Dawei
Yin
[Publish]
The World Wide Web Conference. ACM, 2019.
4. Social Recommendation
[1] Paul Resnick and Hal R Varian. 1997. Recommender systems. Commun. ACM 40,3 (1997), 56–58.
[社会理論]
ユーザーは周辺の人々から情報を集め、影響を受ける [1]
→周りの人との関係性が情報選択に大きな役割を担っている
アイテムの推薦に
social network(人の繋がり)の情報を用いる推薦
- 社会理論に基づいた推薦方法
5. Graph Neural Networks(GNNs)
グラフにおける深層学習技術GNNsが大きく発展
- 周りのノードの構造情報とノードの属性情報を取り込み、ベクトル表現を学習
GCNsの概略 from [2]
[2] T. N. Kipf and M. Welling, “Semi-supervised classification with graph convolutional networks,” in Proceedings of the 6th International Conference on Learning Representations,
2017.
[3] Gilmer, Justin, et al. "Neural message passing for quantum chemistry." Proceedings of the 34th International Conference on Machine Learning-Volume 70. JMLR. org, 2017.
MPNNs式 from [3]
input
Ho(属性行列),
A(隣接行列)
output
Y(クラスラベル行列)