Research On The Construction Of Enterprise Product Network Word Of Mouth Monitoring And Analysis System

With the rapid development of Internet,the emergence of all kinds of network community platforms has provided a good channel for consumers to publish word of mouth information.The Internet is quietly becoming the main position of consumer opinion sharing.At the same time,it is also an important place for enterprises to collect consumers' feedback information,and then find their own product defects,find service blind spots and grasp their own competitive situation in the industry.However,facing the explosive growth of network information,it has been unable to meet the actual needs of word of mouth monitoring and analysis just by relying on traditional manual collection,sorting and analysis.Enterprises urgently need an automated application tool to help them tap the important values of word of mouth information,so as to help improve their products,enhance their competitiveness and find potential risks in negative word of mouth information.Therefore,it is of great practical significance and practical application value to establish an intelligent platform for monitoring and analysis of internet word-of-mouth.Based on the actual needs of enterprises,this paper analyzes the main functional requirements of the application system platform,builds the overall framework of the product oriented online word-of-mouth monitoring and analysis system,and analyzes the key technologies and implementation plan.At the same time,this paper designs the evaluation index system of word-of-mouth monitoring and analysis to provide support for the analysis and visualization of the word-of-mouth.On the basis of the analysis of the system technology system,this paper focuses on the key technologies of the word-of-mouth information acquisition,preprocessing and word of mouth analysis.In the core part of the word-of-mouth analysis,because the traditional sentiment analysis is mostly coarse granularity and ignores the sentence semantics,this paper adopts the method of fine grained opinion mining and emotion analysis based on dependency parsing.The part of word-of-mouth acquisition and preprocessing needs to achieve the acquisition of data as well as the preprocessing operations such as clause participle,part of speech tagging and dependency parsing.In the process of specific word analysis,based on the analysis of syntactic patterns and commentary features of product reviews,this paper summarizes and designs the rules of evaluation unit extraction and emotional calculation in the opinion mining,and carries out an empirical study on the automobile industry.Besides.according to the requirement of word-of-mouth monitoring and analyzing the application system,this paper designs and implements the main functional modules,including the acquisition and analysis of word of mouth,the presentation of results,and the management module of thesaurus and dictionaries.