Search International and National Patent Collections

1. (WO2017149911) DOCUMENT CLASSIFICATION DEVICE, DOCUMENT CLASSIFICATION METHOD, AND DOCUMENT CLASSIFICATION PROGRAM

Pub. No.:    WO/2017/149911    International Application No.:    PCT/JP2016/088160
Publication Date: Sat Sep 09 01:59:59 CEST 2017 International Filing Date: Thu Dec 22 00:59:59 CET 2016
IPC: G06F 17/30
Applicants: RAKUTEN, INC.
楽天株式会社
Inventors: MURAKAMI Koji
村上 浩司
MITA Masato
三田 雅人
Title: DOCUMENT CLASSIFICATION DEVICE, DOCUMENT CLASSIFICATION METHOD, AND DOCUMENT CLASSIFICATION PROGRAM
Abstract:
A document classification device according to one embodiment of the present invention is provided with a generation unit and an updating unit. The generation unit executes first machine learning using, as input data, a subject document to which has been given a correct answer path in a tree structure in which each node indicates a document category, and thereby generates a classification model that indicates a correct path down to a terminal node with regard to the subject document. The updating unit executes second machine learning in which a subject document to which a correct answer path has not been given is applied to the classification model, and, when a path from an N hierarchy node to an N+1 hierarchy node is not the correct answer path, updates the classification model by setting a correction path from the N+1 hierarchy node to an N+2 hierarchy node that is not a child node of the N+1 hierarchy node.