Chatbot de asistencia personal para la recomendación de productos en una tienda de comercio electrónico

Abstract

Abstract: Technology is currently growing exponentially, the great technological changes mean that companies continue to improve and innovate, emphasizing the effectiveness of customer service, since these are the ones who decide to buy or have a service or product.This Degree Project proposes the development of a chatbot for recommending products in an e-commerce store.For the development of the chatbot, different technologies were used, such as the W atson Assistant service, where intentions, entities and assistant dialogue are configured, the integration of a Recommender System was also chosen, based on Machine Learning techniques such as the application of the K algorithm. -NN to generate item recommendations to users appropriately.The performance of the Recommender System was evaluated with three metrics, Precision, Recall and Accuracy, reaching a very good average of 0.84% with the number 11 as K-neighbor, reflecting good performance and quality.Finally, a web application is developed where the operation of the virtual assistant is presented together with the recommendations of items for users.