Skip to Main content Skip to Navigation
Conference papers

Extraction de propriétés de produits

Abstract : In the work presented here, we try to automatically extract some product properties from descriptive texts provided by a merchant website. The constitution of an annotated reference corpus reveals some problems, not only due to the texts but also to the specificities of the task. To handle it, two distinct approaches have been tested : an extraction method based on dictionaries and a machine learning approach making use of CRFs (Conditional Random Fields), for which a large number of models have been tried. The results of our experiments outline the advantages and drawbacks of these two methods
Document type :
Conference papers
Complete list of metadatas

Cited literature [11 references]  Display  Hide  Download
Contributor : Tian Tian <>
Submitted on : Tuesday, February 28, 2017 - 2:45:20 PM
Last modification on : Thursday, April 2, 2020 - 1:28:58 PM
Document(s) archivé(s) le : Monday, May 29, 2017 - 12:14:17 PM


Publisher files allowed on an open archive


  • HAL Id : hal-01473389, version 1



Patrick Marty, Tian Tian, Isabelle Tellier. Extraction de propriétés de produits. COnférence en Recherche d’Information et Applications (CORIA 2014), Mar 2014, Nancy, France. pp.121-136. ⟨hal-01473389⟩



Record views


Files downloads