User Demand Research Based On Online Reviews Data Mining

The existence of products comes from the needs of users.In the rapidly changing market environment,the dynamic process,the user needs often enterprises need according to the market changes and updates in a timely manner to reasonably adjust the direction of enterprise management,is likely to make to the greatest extent to the stages of market development and user demand decision objective,eagerly anticipates the enterprise long-term health development.At present,more and more consumers interact with information and express their emotions through online comments.Online review has become an important channel for enterprises to gain insight into users' needs and intentions in the Internet environment,and also an important way for consumers to obtain information about products and services and reduce information asymmetry.However,the massive amount of online review data creates a huge information overload,making it difficult for businesses and consumers to make rational judgments in a short period of time.Therefore,how to make full use of the online reviews of products to identify and analyze user demand,and reveal the changing trend of user demand,so as to guide the enterprise's product positioning and market competition strategy is a problem worth studying.On the basis of combing existing research theories,this paper firstly uses LDA model to conduct topic mining of Huawei mobile phone online reviews,identify user needs in the results of topic clustering,and establish a user demand factor system.Next,this paper sets up a demand questionnaire according to the basic theory of Kano model,combines the user satisfaction index and the refined Kano model,analyzes the influence of various service elements on user satisfaction,classifies user needs,and determines the importance of various user needs and the order of supply priority.Finally,in this paper,the online reviews divides the time slice mining of time series data of the key subject,at the same time refer to community found the thought of social network analysis,selection of representative range keywords co-occurrence analysis,with the tools of Gephi co-occurrence network mapping and visualization of the subject matter,the changes of user demand analysis of time series.The research results show that when the topic mining is carried out on the three product reviews of the jd.com platform Huawei Changxiang9,P30 and Mate30 5G,the characteristics of the product's own attributes are discussed by users in a stable and high frequency,which is the most important demand that users pay attention to in the process of product purchase and review.The discussion of service quality,accessories and price is relatively low,and the specific topic of each product is slightly different.In the research on the classification of user demand elements,the main demands of smartphone products described by users in the actual comments are divided into charm demand,expectation demand,necessity demand and irrelevant demand.According to the factor characteristics of each category,a series of Suggestions on decision optimization are provided to brand managers.In the time series analysis of online reviews of Mate30 5G products in Huawei mall,it was found that each time segment showed different key theme words,and the representative interval was selected for topic co-occurrence analysis.It was found that the topic fluctuated with time to a certain extent,and the topic popularity was affected by the situation.Then we get the user demand evolution trend.At the theoretical level,this paper integrates the time factor in the process of online comment topic identification,and uses data mining technology to effectively solve the problems such as demand source lag in the traditional user demand identification method,providing a new perspective for the research of online comment user demand mining and analysis.On a practical level,this article research conclusion can effectively guide the consumer purchase decision,at the same time,this article provides the user requirements identification and evolution of the monitoring method can help enterprises to further understand the user key requirements,and timely attention to the new requirements,in addition,for different categories of needs,enterprises can choose according to actual circumstances demand supply priority;Therefore,the results of this study provide an important reference for enterprises to improve their product and service quality and improve user stickiness.