Abstract:

With the popularity of online food trading,rapid increase in user evaluation data,it is more and more necessary to make full use of evaluation information and analysis the potential value in a large number of comments.Traditional sentiment analysis can only identify the overall praise or poor evaluation,can not understand the deep demand of users to make accurate recommendations.In order to solve this problem,a recursive neural network for multiple attribute clustering weighting output model is proposed in this paper.According to attribute words,the model excavates characteristics of the user's interest points.A recommendation algorithm based on user interest points and shop's characteristics.The test results show that the proposed model has a high precision and recall for emotional classification,can accurately capture interest point,can improve the effect of personalized recommendation.

Key words: sentiment analysis, recursive neural network, attribute clustering, deep learning, personality recommendation algorithm