Systematic Review on Online Social Media Recommender Systems

The social network and the social context are two vital elements in social recommender systems. Social recommender systems are an important part of everyday life, on social media. New approaches also need to be adopted to be able to deal with the different needs of social media users. A total of twenty papers were reviewed from the years 2015-2021. There was a peak between the years 2019 and 2020 that shows that research in this area is growing. The results show that RMSE and MAE are the most used evaluation metrics. Precision and recall are the second commonly used evaluation metrics. Natural Language Processing (NLP) classifiers are becoming popular in recommender systems and are used for extracting information from textual analysis to enrich the description within a predictive algorithm. There is not enough research that focuses on online evaluation therefore it is necessary to explore online evaluations and improve results in those experiments.