Abstract

Currently, most recommendation systems do not consider the context in which they are executed, being inappropriate to operate on mobile devices, this can be observed in the field of advertising, where users are overwhelmed by the excessive general information that they receive, causing widespread dissatisfaction with their use. One of the biggest challenges to incorporate contextual information to the software is the design of a formal model for its representation, because traditional methods are inadequate for this purpose, being necessary to use alternative approaches such as those based on ontologies. This work describes the process used in the construction of an ontology to represent the information of the advertisements and the contextual dimensions: location, time and users’ needs, to consider when recommending. Through the application of the NeOn methodology, an expressive and extensible ontological model was obtained that integrates the ontologies: FOAF, OWL-Time and WGS84 Geo Positioning. The proposed ontology is an initial contribution for the creation of a context-aware mobile recommender system of advertisements.