Skip to Main content Skip to Navigation
Patents

Text categorization based on co-classification learning from multilingual corpora

Massih-Reza Amini 1, * Goutte Cyril
* Corresponding author
Abstract : The patent describes a method and a system for generating classifiers from multilingual corpora including subsets of content-equivalent documents written in different languages. When the documents are translations of each other, their classifications must be substantially the same. Embodiments of the invention utilize this similarity in order to enhance the accuracy of the classification in one language based on the classification results in the other language, and vice versa. A system in accordance with the present embodiments implements a method which comprises generating a first classifier from a first subset of the corpora in a first language; generating a second classifier from a second subset of the corpora in a second language; and re-training each of the classifiers on its respective subset based on the classification results of the other classifier, until a training cost between the classification results produced by subsequent iterations reaches a local minima.
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-00871809
Contributor : Massih-Reza Amini <>
Submitted on : Thursday, October 10, 2013 - 3:45:32 PM
Last modification on : Monday, April 20, 2020 - 11:24:01 AM

Identifiers

  • HAL Id : hal-00871809, version 1

Collections

CNRS | LIG | UGA

Citation

Massih-Reza Amini, Goutte Cyril. Text categorization based on co-classification learning from multilingual corpora. United States, Patent n° : 20060101. 2013. ⟨hal-00871809⟩

Share

Metrics

Record views

241