Research On Truck Users’ Requirements Analysis Based On Online Comments

In the context of the new generation of information technology,thanks to the popularization and development of social networks,online comments have become a new carrier for expressing users’ needs and an effective way to obtain product defect information.Through in-depth analysis of online product reviews,users’ needs can be more accurately identified and understood,and the trend of users’ needs over time can be better grasped,which is extremely important for the positioning of a company’s products.The main content of this study is the application of users’ online comments to the process of aiding the optimal design of trucks.By reviewing existing research theories,a truck user demand mining method based on social media online comment texts is proposed.This method extracts truck design attributes and corresponding evaluations from online comments,and then conducts emotional analysis and quantification of the comment texts,and finally constructing a demand prioritization model based on users’ attention and satisfaction with the design attributes.The brief process is as follows:extensively collect the literature and online comments in the field of trucks as the original data set,manually screen out effective comments,and then use natural language processing technology to conduct text preprocessing to build a professional vocabulary-attribute vocabulary,emotional vocabulary,negative vocabulary and degree adverb vocabulary,use six supervised classification models for vocabulary extraction and text classification,use the K-fold cross validation method to calculate the various indicators of the classification model,select the optimal text classification results,conduct emotional analysis and quantification on explicit and implicit sentence structures,ultimately construct a demand priority ranking model based on users’ attention and satisfaction.In response to the shortcomings and shortcomings of current research on users’ demand mining based on comment texts,this study provides a new approach and method for user demand mining for products with complex attribute levels,and mainly improves the comprehensiveness and accuracy of demand mining from the following four aspects:Firstly,cut long sentences into short sentences based on punctuation mark after manually filtering out valid comment texts,and then mark them manually.Secondly,in the stage of text classification and vocabulary extraction,use K-fold cross validation method is used to select the text classification results with the highest F1 from six supervised machine learning and deep learning based text classification models for subsequent research.Thirdly,in the emotional analysis and quantification stage,effective comment texts are divided into three sentence structures-explicit sentence structures containing both attribute and emotion vocabulary,implicit sentence structures containing only attribute vocabulary,and implicit sentence structures containing only emotion vocabulary.Effective methods for extracting implicit needs are established,and emotional analysis and quantification research are conducted on the above three sentence structures.Fourthly,this study proposes a weighting coefficient influenced by environmental factors to weaken the impact of natural conditions on truck users’ satisfaction evaluation of some design attributes.Finally,the importance and need for improvement of truck attributes are ranked based on user attention and emotional values,and a demand priority ranking method that integrates the advantages of multiple models is established.In summary,this study is based on online reviews to explore the diverse needs of truck users.By combining qualitative and quantitative analysis methods,the natural language of users is transformed into product design indicators,which can provide efficient product development planning for design decision-makers in enterprises or relevant departments.