講演抄録/キーワード |
講演名 |
2019-08-30 09:00
ペアマッチング予測のためのFactorization Machinesに基づくニューラルネットワーク ○張 陸瑋・汪 雪婷・山崎俊彦(東大) MVE2019-14 |
抄録 |
(和) |
(まだ登録されていません) |
(英) |
Online dating services have become popular in the modern society. Pair matching prediction between two users in these services can help efficiently increase the property of finding their life partners. Factorization Machines (FMs) can model sparse categorical features, which are effective to many prediction tasks of web applications. To solve the pair matching task, we apply several FM based methods with Deep Neural Networks including DeepFM, NFM and AFM. We also improve them by enhancing the representation of feature interaction embedding and the interaction pooling layer. Through the experiments on a real-world dataset provided by an online dating company, we demonstrate the superior performances of FM-based neural networks. |
キーワード |
(和) |
/ / / / / / / |
(英) |
Pair Matching / Neural Network / Factorization Machines / Recommendation / / / / |
文献情報 |
信学技報, vol. 119, no. 190, MVE2019-14, pp. 49-54, 2019年8月. |
資料番号 |
MVE2019-14 |
発行日 |
2019-08-22 (MVE) |
ISSN |
Print edition: ISSN 0913-5685 Online edition: ISSN 2432-6380 |
著作権に ついて |
技術研究報告に掲載された論文の著作権は電子情報通信学会に帰属します.(許諾番号:10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
PDFダウンロード |
MVE2019-14 |