<![CDATA[Recommender systems (RS) have been widely used to extract relevant and meaningful information from a vast body of data, to make appropriate suggestions to users with different preferences in various domains of applications. However, despite the success of the early recommendation systems, they suffer from two major challenges of cold start and data sparsity. Traditional RS consider an interaction between user and item (2D), neglecting contextual information such as location, until fairly recently. The contexts extend traditional RS to multi-dimension interaction and provides a useful information that allow recommendations to be more personalized. Surprisingly, taking these contexts such as location, into consideration eliminates the challenges of traditional RS. Location-Aware Recommender System (LARS) takes user's location into account as an additional context. The combination allows the prediction of spatial items, items closest to the users, to reduce information overload and was proved to be more effective than earlier RS. In this research, we provide a systematic literature of the existing literature in LARS from 2010 to 2021, focusing on the state-of-the-art methodologies, the domain of applications, and trends of publications in LARS. The paper proposed several models of LARS based on the traditional RS methodologies, providing future directions to researchers. Despite numerous reviews available on LARS, a review that proposed several LARS techniques were not found in the literature. The results indicated that the trend of publication in LARS is growing exponentially and that the field is getting attention rapidly with the number of publications on the rise every year.]]>
Copyrights © 2021
