Diseño de un sistema de recomendaciones de paquetes prepago mediante uso de técnicas de inteligencia artificial en redes telefónicas en Ecuador.

Abstract

Technological evolution throughout the history of telecommunications companies has generated new ways of communicating, which has led to the generation of disruptive services in the industry. We must consider the current reality where there is an overwhelming increase in the use of data by users and the contribution to the economy with 5% of GDP (in Latin America), in addition to the positive impact on jobs and technological services. This work "Designing a system of recommendations for prepaid packages using artificial intelligence techniques in telephone networks in Ecuador" aims to recommend prepaid packages to users of the various telephone operators. The methodology applied in this work is the life cycle of data science to develop the problem, model to be used and the metric evaluated to obtain it’s precision rating; recommendation systems with a collaborative filtering approach; the Lorenz curve and the Gini coefficient to determine the distribution of customer income. The Jupyter Notebook tool and the Python programming language are used to apply the methodology of the proposed system which is composed of the statistical analysis of the data obtained, the primary variables generated from the interviews with collaborators of the telecommunications companies, the algorithm recommendation of packages for each user and the evaluation of the accuracy of the results. The results obtained by the system led to the conclusion that the collaborative filtering approach, the variables used in the system, and the use of the cosine similarity metric allow the recommendation of the packages to the users with a precision greater than 60 %.