Skip to Main content Skip to Navigation
Conference papers

Extraction générique de connaissances à partir de données textuelles et mesure de la performance des systèmes d’extraction de relations dans un contexte non supervisé.

Abstract : Among the incoming challenges in the industrial domain and in the monitoring of industrial systems, the aggregation, synthesis and management of knowledge through ontological structures occupy an essential place. Existing knowledge extraction systems often use a supervised approach which rely on labelled data for which the annotation process is fastidious. This paper presents an unsupervised self-feeding rule-based approach for domain-independent ontology population from textual data. Moreover, the evaluation of such systems, performing knowledge extraction using natural language processing methods requires the use of performance indicators. The indicators usually used in such evaluations have limitations in the specific context of knowledge extraction for unsupervised ontology population. Thus, the definition of new evaluation methods becomes a need arising from the singularity of the harvested data, especially when these are unlabelled. Hence, this article also proposes a method for measuring performance in a context where reference data and extracted data do not overlap optimally. The proposed evaluation method is based on the exploitation of data that serve as a reference but are not specifically linked to the data used for extraction, which makes it an original evaluation method.
Document type :
Conference papers
Complete list of metadata

https://hal-mines-albi.archives-ouvertes.fr/hal-03331800
Contributor : Imt Mines Albi Imt Mines Albi <>
Submitted on : Thursday, September 2, 2021 - 10:38:47 AM
Last modification on : Sunday, September 5, 2021 - 3:05:36 AM

File

Extraction-generique-de-connai...
Publisher files allowed on an open archive

Identifiers

  • HAL Id : hal-03331800, version 1

Citation

Yohann Chasseray, Anne-Marie Barthe-Delanoë, Jean-Marc Le Lann, Stéphane Negny. Extraction générique de connaissances à partir de données textuelles et mesure de la performance des systèmes d’extraction de relations dans un contexte non supervisé.. CIGI-Qualita21 : 14ème Conférence Internationale Génie Industriel QUALITA, May 2021, Grenoble (à distance), France. pp.660-668. ⟨hal-03331800⟩

Share

Metrics

Record views

19

Files downloads

11