Global Patent Index - EP 1210669 A1

EP 1210669 A1 20020605 - DOCUMENT CLASSIFICATION APPARATUS

Title (en)

DOCUMENT CLASSIFICATION APPARATUS

Title (de)

DOKUMENT-KLASSIFIKATIONS-VORRICHTUNG

Title (fr)

DISPOSITIF DE CLASSIFICATION DE DOCUMENTS

Publication

EP 1210669 A1 20020605 (EN)

Application

EP 99946569 A 19990825

Priority

SG 9900089 W 19990825

Abstract (en)

[origin: WO0114992A1] Document classification apparatus is disclosed comprising feature extraction means (20) for extracting a plurality of features from a document and a classifier (30) operable to process the document in a knowledge acquisition mode in which information for classification is acquired from the features of the document and added incrementally to a knowledge base (105) and in a document classification mode in which the classifier (30), using the knowledge base, determines from the features a predicted classification for the document, the classifier (20) being switchable between the modes under user control. The apparatus is further operable to perform rule insertion in the knowledge acquisition mode in which a number of features are input by a user to the classifier together with a classification with which the features are associated.

IPC 1-7

G06F 15/80; G06K 9/66

IPC 8 full level

G06F 17/30 (2006.01); G06K 9/20 (2006.01); G06K 9/66 (2006.01); G06N 5/02 (2006.01)

CPC (source: EP)

G06F 16/353 (2019.01); G06N 5/025 (2013.01)

Citation (examination)

  • AH-HWEE T: "Adaptive Resonance Associative Map", NEURAL NETWORKS, ELSEVIER SCIENCE PUBLISHERS, BARKING, GB, vol. 8, no. 3, 1995, pages 437 - 446, XP004011586
  • DUMAIS S. AND PLATT J. AND HECKERMAN D. AND SAHAMI M.: "Inductive Learning Algorithms and Representations for Text Categorization", PROCEEDINGS OF THE 1998 ACM CIKM INTERNATIONAL CONFERENCE ON INFORMATION AND KNOWLEDGE MANAGEMENT, BETHESDA, MARYLAND, USA, 3 November 1998 (1998-11-03)
  • See also references of WO 0114992A1

Designated contracting state (EPC)

DE FR GB

DOCDB simple family (publication)

WO 0114992 A1 20010301; EP 1210669 A1 20020605

DOCDB simple family (application)

SG 9900089 W 19990825; EP 99946569 A 19990825