To provide means for calculating a potential feature amount in which a tendency of a user regarding how to give an expression is considered.
An information processing device expresses a variable c representing a user cluster with a random variable following a distribution different by each user, expresses a variable y representing a topic on content items with a random variable following a distribution different by each content, expresses a variable x representing an expression cluster with a random variable following a distribution different by each combination of a user cluster and a topic, calculates, based on a probability model in which a variable a representing an expression of a user is expressed with a random variable following a distribution different by each expression cluster, a parameter included in the probability model using a combination of a feature amount of content and an expression given as learning data, and calculates an attribution degree of a user to a user cluster and an attribution degree of an expression to an expression cluster using the parameter.
| JP2009157907A | 2009-07-16 | |||
| JP2010237864A | 2010-10-21 | |||
| JP2010108265A | 2010-05-13 | |||
| JP2009157907A | 2009-07-16 | |||
| JP2010237864A | 2010-10-21 | |||
| JP2010108265A | 2010-05-13 |
| US20090144226A1 | 2009-06-04 | |||
| WO2007066607A1 | 2007-06-14 | |||
| US20090144226A1 | 2009-06-04 | |||
| WO2007066607A1 | 2007-06-14 |
CSNJ201310057175; 古賀 裕之 他: '潜在トピックに着目したＴｗｉｔｔｅｒ上のユーザ推薦システムの構築' ヒューマンインタフェースシンポジウム２０１０ 論文集 Ｎｏ．３２４３, 20100910, ｐｐ．８６７-８７２．, ヒューマンインタフェース学会
JPN6014026255; 古賀 裕之 他: '潜在トピックに着目したＴｗｉｔｔｅｒ上のユーザ推薦システムの構築' ヒューマンインタフェースシンポジウム２０１０ 論文集 Ｎｏ．３２４３, 20100910, ｐｐ．８６７-８７２．, ヒューマンインタフェース学会
Tetsuo Kanamoto
Koji Hagiwara
Kazuki Matsumoto
Next Patent: DIGITAL MOVING IMAGE ACCESS CONTROL SYSTEM AND PROGRAM