Extraction de contextes de citations dans un corpus de publications scientifiques

Abstract : To evaluate their activities, researchers have to prove that their works have a real impact on the scientific community. The dedicated discipline is called “bibliometrics”. However, most bibliometic tools count quotation of an article with different criteria. There do not take into account the opinion, positive or negative, of their readers. We studied the possibility to offer an application providing a qualitative analysis in order to have a better idea of the effective influence of scientific production. We based ourselves on semantic and syntactic analysis. We proposed a weighting of this elements as to obtain an evaluation model, combining quantitative and qualitative analysis of texts. Finally, we have evaluated robustness of our program and considered other developments and better data optimization.
Document type :
Master Thesis
Complete list of metadatas

https://dumas.ccsd.cnrs.fr/dumas-01666235
Contributor : Uga - Bulles <>
Submitted on : Monday, December 18, 2017 - 10:48:26 AM
Last modification on : Tuesday, April 2, 2019 - 1:46:36 AM

File

mémoire_verdenet_I3L.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : dumas-01666235, version 1

Citation

Valérie Verdenet. Extraction de contextes de citations dans un corpus de publications scientifiques. Sciences de l'Homme et Société. 2017. ⟨dumas-01666235⟩

Share

Metrics

Record views

249

Files downloads

89