Exploiting geo-referenced data as contextual information into context-aware recommender systems

• Published in: 2017 36th International Conference of the Chilean Computer Science Society (SCCC) pp. 1 - 7
• Main Authors: Domingues, Marcos Aurelio, Pimentel, Maria da Graca Campos
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.10.2017
• Subjects: Clustering algorithms; Collaboration; Context modeling; Context-Aware Recommender Systems; Contextual Information; DBSCAN Clustering Algorithm; Filtering algorithms; Geo- Referenced Data; Measurement; Proposals; Recommender systems
• DOI: 10.1109/SCCC.2017.8405110

A recommender system can be defined as an information filtering technology which can be used to output a ranking of items (e.g., products, places, movies, etc) that are likely to be of interest to a user. Thus, such systems can be used in various fields to recommend items of interest to the users. Most recommender approaches focus only on the relation of users and items to make the recommendations. However, in many applications, it is also important to incorporate contextual information into the recommendation process. Context-aware recommender systems makes recommendations by incorporating available contextual information into the recommendation process. Although the use of contextual information in recommender systems has received great focus in recent years, there is a lack of automatic methods to obtain such information for context-aware recommender systems, specially in geo-referenced systems. These systems can generate hundreds of data corresponding to the user coordinates. However, this huge number of observations is not useful if the objective is to represent where the user has been. In such a case, only a single data point is needed to represent the geographical location of the place that the user has visited. In order to obtain such a single data point to represent the user location, we have proposed to use the DBSCAN clustering algorithm to convert the geo- referenced data into data regions (i.e., data clusters). Then, we can use the regions as contextual information into context-aware recommender systems. We have evaluated our proposal in two data sets, which showed evidences that the clustering of geo- referenced data into regions can provide better recommendations.