Research On Recommendation Method Based On User's Opinion Mining In Reviews

As an important solution to the problem of “Information Overload” in the era of big data,recommender system has been widely concerned by academia and industry.With the continuous development of recommender systems,recommendation methods have been widely studied.Traditional methods usually predict users' satisfaction based on the relationships between users and items,but other information which may affect users' satisfaction has not been fully mined and utilized,resulting in low accuracy.Users often express their fine-grained opinion through reviews.Therefore,it is worthy of further exploration and research to fully mining user's opinion in reviews to improve the effect of recommendation methods.Users often express their opinions on multi-aspects of items in reviews.Therefore,a more personalized recommendation method based on mining users' aspect-based opinion in reviews is proposed.The keyword sets of different aspects are built by the technique of text feature extraction,and the aspect expressions are extracted.Then Aspect opinion expressed in reviews is extracted by calculating the sentiment rating of the aspect expression.Based on the representation ability for multi-dimensional features of tensor,and the mining ability for multi-relationship between these features of tensor decomposition,a rating tensor which contains users' aspect-based opinions in reviews is established.Then the rating predicted model of users for items is established through tensor decomposition technique.At last,recommendation is generated through the predicted ratings.Experiments are implemented to evaluate the proposed method,and compared with several other baseline methods at the same time.The results show that the proposed method can not only generate effective recommendation for users but also have a higher rating prediction accuracy compared with several other baseline methods.Furthermore,the proposed method has a better performance when confronted with the problem of data sparsity.