Integrating Sentiment Analysis in Recommender Systems  
Author Thanh Hung Bui

 
Co-Author(s)

 
Abstract Customer product reviews play an important role in the customer’s decision to purchase a product or use a service. Providing a useful suggestion of products to online users to increase their consumption on websites is the goal of many companies nowadays. In this paper, we propose a recommender system based on sentiment analysis. The system is built by integrating sentiment analysis to recommender system in order to generate the most accurate. We use hybrid deep learning method CNN-LSTM for sentiment analysis based on vector of words in the customer product reviews. The result in the sentiment analysis is used to combine the neighbor’s item ratings to produce a prediction value for the target user. This helps the recommender system to generate efficient recommendations for that user. We do experiment in Amazon food review dataset. The proposed model shows interesting results on the impact of integrating sentiment analysis in the recommender systems.

 
Keywords Sentiment Analysis, Word Embeddings, CNN-LSTM, Matrix Factorization, Recommender System
   
    Article #:  DSBFI19-50
 
Proceedings of ISSAT International Conference on Data Science in Business, Finance and Industry
July 3-5, 2019 - Da Nang, Vietnam