Exploitation de différentes approches pour détecter et catégoriser le risque chimique et bactériologique

Abstract : Exploitation of various approaches for automatic detection and categorization of chemical risk Chemical risk corresponds to situations in which chemical products are or can be dangerous for human or animal health, or for environment. Detection of information on risk of chemicals occupies an important place in environmental agencies and researchers. Yet, large amounts of available data and controversies make it difficult to find the relevant information quickly and efficiently. Our objective is to propose an automatic help for the analysis of scientific literature in order to detect sentences indicative of chemical or bacteriological risk. We tackle the task as categorization problem: the sentences are are to be categorized in classes of risk. We use three approaches: rule-based, supervized categorization and information retrieval. The best results are obtained with supervized categorization and information retrieval. According to approaches, the results show up to 0.8 F-measure. MOTS-CLÉS : Risque chimique, catégorisation supervisée, recherche d'information.
Liste complète des métadonnées

Cited literature [13 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01426821
Contributor : Natalia Grabar <>
Submitted on : Wednesday, January 4, 2017 - 10:59:10 PM
Last modification on : Monday, March 18, 2019 - 4:25:05 PM
Document(s) archivé(s) le : Wednesday, April 5, 2017 - 3:31:14 PM

File

grabar-TALN2016RISKchim.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01426821, version 1

Citation

Natalia Grabar, Thierry Hamon. Exploitation de différentes approches pour détecter et catégoriser le risque chimique et bactériologique. Risque et TAL, TALN 2016 workshop, Jul 2016, Paris, France. ⟨hal-01426821⟩

Share

Metrics

Record views

136

Files downloads

147