Sistemas de recomendação que utilizam dados espaciais: um mapeamento sistemático

Abstract

The increasing availability of spatial data, driven by greater connectivity and the massive use of mobile devices, has transformed several areas of technology, especially Recommender Systems. These systems, which already play a crucial role in personalizing user experiences on digital platforms, benefit significantly from the integration of spatial data, leading to more accurate recommendations. However, there is no comprehensive mapping of how this data is utilized in these systems. This dissertation addresses this gap by conducting a systematic review of the incorporation of spatial data in Recommender Systems. The adopted methodological approach is theoretical and exploratory, based on bibliographic and documentary research. The process consists of well-defined stages: planning, execution, and presentation of results. data organization and systematization are performed through systematic tabulation. The main objective is to map how spatial data is being utilized in Recommender Systems. The key findings include: (1) Recommender Systems leveraging spatial data apply Machine Learning and Collaborative Filtering techniques to enhance recommendation accuracy and relevance; (2) the integration of spatial data involves a complex process of collection, extraction, and mapping, typically beginning with the acquisition of geospatial information, such as latitude and longitude coordinates, often derived from Social Media; and (3) baseline models are frequently used to evaluate Recommender Systems that incorporate spatial data.