Multi-Agent System Used for Recommendation of Historical and Cultural Memories

Abstract

In this document the proposal of a recommendation system based on multi agent is made allowing the analysis of user behavior when visiting historical and cultural memories, giving recommendations based on qualifications and duration times for the observation of art pieces. It is also possible to see the system architecture, the server used for the development of the multiagent system, as well as the communication between agents to carry out a route, and the functionality for recommending new routes to a user. The multi-agent system uses a neural network that allows to analyze the behavior of a user in a route; using the feedback given for the neural network the data is checked, allowing determine the user preferences. A set of historical and cultural memory data set is used to generate recommendations; in addition, a user storage API is employed. For the system visualization, this prototype is connected with an augmented reality application that allows users access to visit art pieces and use predefined preferences.