A Job Recommender System for the Unique Framework of Classification of Occupations in Colombia (CUOC) Via Collaborative Filtering

Abstract

Objective: The objective of this research is to improve the match between labor supply and demand in Colombia by using machine learning techniques and the Unique Framework of Classification of Occupations in Colombia (CUOC). This framework allows us to enhance the alignment between resumes and job offers, helping job seekers obtain personalized job recommendations based on their profiles. Materials and Methods: This proposal uses a combination of clustering, classification, and collaborative filtering algorithms to obtain the ten best available vacancies for a particular resume. Standardization of resumes and job offers was performed during the preprocessing stage. We utilized natural language processing algorithms to extract attributes from the CUOC framework. For the training process, we initially employed the K-means algorithm to group the attributes of the CUOC framework. Therea fter, we used KNN, DNN, and AdaBoost as classification algorithms to develop a model that best correlates a resume with a group of vacancies. Finally, a web application was developed using the Django framework, providing a user-friendly interface for job seekers to receive recommendations on the basis of the model outcomes. Results and Discussion: The best model was selected on the basis of accuracy and processing time. The results indicate that the highest accuracy and recommendation performance were achieved via the CUOC framework, generating a recommendation of the top 10 vacancies on the basis of their similarity level. Conclusion: Including CUOC characteristics in both resumes and vacancies allows for better matching within the context of the Colombian labor market.