Information processing for mood-based contextual recommendation in mobility environments

Context-awareness is a technological trend that adds value in many fields. In this work, a research on context that is associated to public information contents is presented. The global objective is to perform fully contextual recommendations in a variety of scenarios, where the context model extracted for information items is used as the basis for the enhanced recommendation. Specific works provided on this thesis are the details about the mechanisms to extract context from user-generated contents. Such mechanisms are based on a combination of text mining algorithms, a flexible model and execution architecture that enables the processing of any kind of item for the purpose of context extraction. Apart of the processing model details, a complete prototype is implemented and presented as a "Telefonica I+D" project called Walkopedia?, for a specific use case, especially interesting for telecommunications operators, as context-based recommendations in mobility environments, where getting information highly-personalized not only for the specific user but also to specific context is a must