ATTENTION/WARNING - NE PAS DÉPOSER ICI/DO NOT SUBMIT HERE

Ceci est la version de TEST de DIAL.mem. Veuillez ne pas soumettre votre mémoire sur ce site mais bien à l'URL suivante: 'https://thesis.dial.uclouvain.be'.
This is the TEST version of DIAL.mem. Please use the following URL to submit your master thesis: 'https://thesis.dial.uclouvain.be'.
 

NLP for recruitment

Ewel, Florence ; Taburiaux, Alexandre
(2024)

Files

Ewel_50111700_Taburiaux_11861900_2024.pdf
  • Closed access
  • Adobe PDF
  • 3.66 MB
Ewel_50111700_Taburiaux_11861900_2024_Errata.pdf
  • Closed access
  • Adobe PDF
  • 4.49 MB

Details

Supervisors
Jodogne, Sébastien
Faculty
Ecole polytechnique de Louvain
Degree label
Master [120] en sciences informatiques, à finalité spécialisée
Abstract
Today, job offers are published directly on the websites of the companies and/or on dedicated websites. Candidates apply directly through the website and send their cover letter and curriculum vitae electronically, usually in PDF format. The processing of these applications is a time-consuming process and requires a lot of expertise and knowledge about the position to be filled. The greater the number of applications received, the more difficult the task will be. In the IT sector, job offers and applications (covering letters and curricula vitae) are written in French or in English. As part of our Master's program, we decided to develop an artificial intelligence system to classify CVs according to their correlation with the job offer. An artificial intelligence with this objective must meet a number of constraints. Firstly, it must be able to read PDF formats and recognise the language in which CVs are written. Secondly, it must allow CVs to be classified in a way that is relevant to the vacancy. It also has to meet practical requirements, such as being able to generate the filing in a reasonable amount of time and, ideally, using few IT resources, so it's easy to use. Finally, in order to be functional, the presentation of the application has to be easy to understand and easy to use. We have created a tool for reading PDFs and determining the language in which they are written. Next, we researched artificial intelligence models (Natural Language Processing). We then adapted these models for our application. We tested these different tools on a database consisting of 2 sets of 6 job offers and 6 to 10 associated CVs, in 2 languages: French and English. This allowed us to objectively assess the performance of each model in each language. For the best-performing models, we assessed the IT resources required to implement them and the time required to achieve results. All of these elements have enabled us to implement an artificial intelligence that meets our initial objective: to develop an artificial intelligence system that classifies CVs according to their correlation with the job offer and responds to the material constraints of time management and IT resources.