Aplicação de regras de associação para o desenvolvimento de sistemas de recomendação de materiais bibliográficos da UFRN

Abstract

The objective of this work is to develop a system for recommending books from the regular collection of the UFRN library network. The recommendation system used data from loans made in previous semesters and for the generation of association rules, the algorithm FP-Growth was used to indicate possible references that can be used together, in a way that facilitates the assimilation knowledge on the part of the student. The methodology used is composed of the following steps: exploratory data analysis, training, validation of machine learning models and data visualization. For the development of the solution, the set of data analysis libraries of the Python language was used. For the validation of the proposal, information on book loans in UFRN libraries was used, obtained from the institution’s open data portal. In terms of results, three combinations of variables were analyzed that generated between ten and forty association rules.