Title:
DOCUMENT CLASSIFICATION DEVICE AND TRAINED MODEL
Document Type and Number:
WIPO Patent Application WO/2020/021845
Kind Code:
A1
Abstract:
A document classification device for generating, through machine learning, a document classification model, which is a model for classifying documents and which, on the basis of an input document, outputs identification information identifying a classification result, said document classification device being provided with: an acquisition unit which acquires learning data including a document and identification information associated with the document; a feature quantity extraction unit which extracts, as feature quantities, words included in the document and character information comprising one or a plurality of items of information extractable from the words, the character information comprising one of the characters constituting a word or a plurality of successive characters in the word; and a model generation unit which performs machine learning on the basis of the feature quantities extracted from the document and the identification information associated with the document, and generates a document classification model.
Inventors:
IKEDA TAISHI (JP)
Application Number:
PCT/JP2019/021289
Publication Date:
January 30, 2020
Filing Date:
May 29, 2019
Export Citation:
Assignee:
NTT DOCOMO INC (JP)
International Classes:
G06F16/35; G06V30/40
Foreign References:
US20060089924A1 | 2006-04-27 | |||
CN108108351A | 2018-06-01 | |||
JPH08166965A | 1996-06-25 | |||
JPH0554037A | 1993-03-05 |
Attorney, Agent or Firm:
HASEGAWA Yoshiki et al. (JP)
Download PDF: