Extraction and Use of Structured and Unstructured Features for the Recommendation of Urban Resources

Brenda Salenave Santana, Leandro Krug Wives

Abstract: Urban Computing is concerned about the exploration and understanding of urban systems using data generated by itself. The objective of this paper is to describe an approach to analyze information expressed in social networks to help the recommendation of urban resources. This process considers different structured and unstructured features like resource’s location, reviews polarity, and user profile reliability. Therefore, we use text and Web mining techniques to extract those features and then apply traditional recommendation algorithms considering different combinations to identify if they provide better results. Results were compared, and we found that for neighborhood algorithms, the proposed approach presented better results when compared to traditional methods.