Research On The Method Of User Groups' Key Requirements Mining With Fine-grained Sentiment Analysis

With the rapid development of the Internet and Electronic Commerce,more and more customers tend to comment on what they bought online,which results in a large number of user-generated content(UGC).Meanwhile the UGCs have a great impact on the decision of the later buyers.Whether the reviews is positive or negative,those are reflecting the users' attitude to the products.It reminds that mining customers' reviews has commercial value on product development and publicity.Therefore,the target of this paper is to mine the users' key requirement from the sellers,in order to achieve more wealth.The method we used is the fine-grained sentiment analysis.Based on content data,this paper firstly crawls and preprocesses the Chinese online comments data,then establishs a double propagation algorithm to extract the product aspect and opinion words.Secondly,by calculating the spatial similarity of words,the k-means algorithm is used to classify the original attribute words.Thirdly,based on the classification of opinion word,determine the polarity of these words and the inclication of aspect words.Furthermore,in order to avoid the shortcomings of the user demand analysis results,this paper takes the product aspects as the node and the co-occurrence of aspects as the side to construct the network model.Then uses the network clustering analysis method to extract the core aspects.Finally,summarize the results of the analysis,to provide business decision-making recommendations.Some contribution as follow.(1)Based on the double propagation algorithm,the fine-grained sentiment analysis performs well even without labelled data set and can be applied to different fields.The product aspect and opinion words extraction are more effective than the traditional methods.(2)In this paper,mobile phone reviews on tmall.com are taken as the dataset,found that the severe competition between mobile phone manufacturers and the consumer demand for mobile products.(3)Taking into account the importance of product aspects and emotional tendencies,the experiment on this paper help manufacturers mining the core user requirement.So that sellers can know better about its user group and the products'merits and drawbacks,so as to provide strategic basis for future products development.