Research On Analysis Method Of Automotive Products Functional Requirements Based On User Review Data

User requirements in the information age tend to be personalized,diversified,and dynamic.How to accurately grasp user requirements and quickly realize their conversion to engineering languages has become a key issue for enterprises.Today,more and more vertical websites,e-commerce websites,product forums,social medias,and Q&A communities can provide consumers with a broad platform for expressing opinions,and the resulting comment content can more truly reflect user requirements and product defects.Mining user requirements based on internet review data and turning them into technical elements that guide product design can help companies better understand the user experience and drive product development faster.This article conducts demand mining on user reviews of automotive vertical websites,and aims to propose a method for analyzing the functional requirements of automotive products based on user review data.Its main research content includes the following aspects:(1)Completed the usefulness screening of user comment data and text sentiment classification.Use Word2 vec to expand the attribute words of the professional lexicon in the automotive field,and supplement the emotional words of the basic sentiment dictionary through semantic orientation pointwise mutual information(SO-PMI)to build an attribute lexicon and sentiment dictionary for the automotive field,and use both to filter useful comments.Based on the bi-directional long short-term memory neural network(Bi LSTM),convolutional neural network(CNN)and Bi LSTM-CNN to perform the performance test of the text classification model.The experimental results show that the Bi LSTM-CNN model has the best text classification effect,and this model can help to identify the emotional polarity of useful comments.(2)This paper proposes a method for analyzing the user requirements of automobile products based on XGBoost.According to the text sentiment classification results of user review data,an analysis model of influencing factors of user satisfaction based on XGBoost is established.The user pain point index is measured in combination with the importance of key influencing factors and their negative emotional intensity,and the improvement direction of automobile products is determined according to the user pain point index ranking.Negative viewpoints are extracted for the car product attributes that need to be iterated,and then the user requirements of car products are analyzed.(3)A method for determining the comprehensive importance of the functional characteristics of automotive products based on the rough house of quality(HOQ)is proposed.Based on user review data,the user requirements of automobile products are obtained,and the integrated method of rough number and best-worst method(BWM)is used to calculate the basic importance of user requirements,and it is revised according to market competitiveness.The basic importance of functional characteristics is calculated based on the comprehensive importance of user requirements and the rough house of quality(HOQ),and the basic importance is revised based on the autocorrelation relationship of the functional characteristics and the technical competitiveness of the functional characteristics,and the comprehensive importances of functional characteristics are determined according to the linear weighted combination method,in order to clarify the focus of automotive product development.