datasets

This project explores the integration of Geographic Information Systems (GIS) and Natural Language Processing (NLP) to improve job–candidate matching in recruitment. Traditional AI-based e-recruitment systems often ignore geographic constraints. Our hybrid model addresses this gap by incorporating both textual similarity and spatial relevance in matching candidates to job postings.

Data Used

Candidate Data (CVs)

• Source: Scraped from emploi.ma
• Size: 1000 CVs after cleaning
• Content: Candidate names (anonymized), skills, experiences, locations (coordinates), availability, etc.

Job Descriptions

• Source: Publicly available dataset from Kaggle
• Size: we took 1000 job postings using category :Morocco
• Content: Titles, descriptions, required skills, sector labels, and office locations...

All datasets have been cleaned and anonymized for privacy and research ethics compliance.