Sentiment-based Movie Recommender System using Deep Learning
Abstract
The emergence of overloaded internet content poses many new challenges for users and content providers alike. To limit the amount of content viewed, such as videos, music, or other products of content providers such as Netflix or Amazon, the recommender systems are used to guide the user via the available material. These services collect knowledge about the customer and strive to deliver personalized experience. Many sophisticated programs have an approach to contents, but often neglect to take into consideration the nature of user sentiments. This leads to the creation of a sentiment-as-input paradigm incorporating current studies on the recognition of human sentiment and emotionally classifying material within the movie domain. Multiple models of the learning are tested and an ANN is selected due to its outstanding performance. The results of this analysis show that the user sentiment should be used to suggest tailor-made content rather than random content.
Authors
Naveed Jhamat
Assistant Professor, Department of Information Technology, The University of the Punjab, Gujranwala Campus, Punjab, Pakistan
Ghulam Mustafa
Assistant Professor, Department of Information Technology, The University of the Punjab, Gujranwala Campus, Punjab, Pakistan
Zeeshan Arshad
Lecturer, Department of Information Technology, The University of the Punjab, Gujranwala Campus, Punjab, Pakistan
Keywords
Artificial Neural Networks, Emotions, Movies Recommendations Recommender Systems