Evaluación del desempeño computacional de sistemas de recomendación aplicado a bases de datos farmacológicas : implementación de un sistema de recomendación de filtrado colaborativo para predicción de interacciones farmacológicas.

Abstract

In recent years with the advances and development of new technologies, a wide variety of smart devices have been born and with it the growth of information on the web which has generated a problem of information overload, which is why they have been developed to over the years recommender systems that allow easy retrieval and classification of information. For a recommendation system to achieve its objective of making recommendations to its users, it needs to collect information from users through a collection and/or feedback process, this information collection is achieved through the use of surveys or article qualifications that allow to classify the information according to the preference of the users.Information overload occurs when there is a large volume of information that can exceed the processing of human beings, which can sometimes generate frustration and/or obtaining irrelevant information. In the pharmaceutical environment, information overload is not an isolated problem, which implies a better treatment of information in order to provide users with recommendations based on their needs and their profile. Hence the importance and growing interest of the technologies involved in these fields of analysis of large amounts of data such as Big Data. The present work focuses specifically on the development of a collaborative filtering optimization system, based on the existing activity between compounds and cells of the pharmacological databases prior to data processing, in this area it seeks to establish a system that allows us to determine the existing activity automatically with the minimum effort required.