Title:
CLUSTERING SYSTEM, METHOD, PROGRAM, AND RECOMMENDATION SYSTEM
Document Type and Number:
WIPO Patent Application WO/2018/078761
Kind Code:
A1
Abstract:
The purpose of the present invention is to provide a clustering system capable of clustering a plurality of types of matters such that a matter in which corresponding character data exists, but relational data with respect to a matter of a different type does not exist, can be recommended as a different type matter. A first clustering means 3001 clusters first IDs on the basis of relational data. A second clustering means 3002 clusters second IDs on the basis of the relational data and character data associated with the second IDs. A topic allocating means 3003 allocates a topic for each word included in the character data associated with the corresponding second IDs. A parameter determining means 3004 determines a usage parameter in a first clustering process, a usage parameter for a second clustering process, and a usage parameter for a topic allocating process. Until a predetermined condition is determined to be satisfied, the aforementioned processes are repeated.
Inventors:
TOMOBE KATSUFUMI (JP)
OYAMADA MASAFUMI (JP)
NAKADAI SHINJI (JP)
OYAMADA MASAFUMI (JP)
NAKADAI SHINJI (JP)
Application Number:
PCT/JP2016/081873
Publication Date:
May 03, 2018
Filing Date:
October 27, 2016
Export Citation:
Assignee:
NEC CORP (JP)
International Classes:
G06F17/30
Foreign References:
| US20090300547A1 | 2009-12-03 | |||
| US20160092781A1 | 2016-03-31 |
Other References:
WANG, CHONG ET AL.: "Collaborative Topic Modeling for Recommending Scientific Articles", PROCEEDINGS OF THE 17TH ACM SIGKDD INTERNATIONAL CONFERENCE ON KNOWLEDGE DISCOVERY AND DATA MINING, 21 August 2011 (2011-08-21), pages 448 - 456, XP055500553
Attorney, Agent or Firm:
IWAKABE Fuyuki et al. (JP)
Download PDF:
Previous Patent: NONAQUEOUS ELECTROLYTE SECONDARY BATTERY
Next Patent: ELEVATOR CAR UPPER HANDRAIL DEVICE
Next Patent: ELEVATOR CAR UPPER HANDRAIL DEVICE