Presentation of a Recommender System with Ensemble Learning and Graph Embedding: A Case on MovieLens

Information technology has spread widely, and extraordinarily large amounts\nof data have been made accessible to users, which has made it challenging to\nselect data that are in accordance with user needs. For the resolution of the\nabove issue, recommender systems have emerged, which much help users go through\nthe process of decision-making and selecting relevant data. A recommender\nsystem predicts users behavior to be capable of detecting their interests and\nneeds, and it often uses the classification technique for this purpose. It may\nnot be sufficiently accurate to employ individual classification, where not all\ncases can be examined, which makes the method inappropriate to specific\nproblems. In this research, group classification and the ensemble learning\ntechnique were used for increasing prediction accuracy in recommender systems.\nAnother issue that is raised here concerns user analysis. Given the large size\nof the data and a large number of users, the process of user needs analysis and\nprediction (using a graph in most cases, representing the relations between\nusers and their selected items) is complicated and cumbersome in recommender\nsystems. Graph embedding was also proposed for resolution of this issue, where\nall or part of user behavior can be simulated through the generation of several\nvectors, resolving the problem of user behavior analysis to a large extent\nwhile maintaining high efficiency. In this research, individuals most similar\nto the target user were classified using ensemble learning, fuzzy rules, and\nthe decision tree, and relevant recommendations were then made to each user\nwith a heterogeneous knowledge graph and embedding vectors. This study was\nperformed on the MovieLens datasets, and the obtained results indicated the\nhigh efficiency of the presented method.\n