|
 
|
[SSL10] e-Recruitment: Detection of Relevant Keywords in the Job Offer Title
Atelier, Poster ou Démonstration dans une Conférence Internationale : JADT'2010, 10th International Conference on Statistical Analysis of Textual Data, Rome, June 2010, pp.xx,
motcle:
Résumé:
The increase in the number of different ways to recruit online has made the issue of job posting performance and its optimization increasingly important. In particular, performance can be assessed by the amount of applications received. In this paper, the goal is to define a strategy for detecting key-words within a job postingÂ’s title that provide information to explain the postingÂ’s performance. We will work on a corpus of job postings characterized by the underlying function (e.g. marketing, finance,Â…). After a pre-processing of filtering on the corpus, we will compute characteristic forms and identify the most frequent forms. In order to confirm the relevance of the key-words obtained, the latters are coded into dichotomic variables and introduced as predictors in a regression tree, in addition to function predictors. The results are encouraging and confirm key-words contribution into explaining performance.
Equipe:
msdma
BibTeX
| @inproceedings { | |||
| SSL10, | |||
| title | = | "{e-Recruitment: Detection of Relevant Keywords in the Job Offer Title}", | |
| author | = | " J. Séguéla and G. Saporta and S. Le Viet ", | |
| booktitle | = | "{JADT'2010, 10th International Conference on Statistical Analysis of Textual Data, Rome}", | |
| year | = | 2010, | |
| month | = | "June", | |
| pages | = | "xx", | |
| } | |||
