The Design And Implementation Of Complaints Customer Behavior Intelligence Analysis System Based On Text Classification

China Mobile in complaint handling process,accumulate large amount of complaints.The contact and interaction between customer and company contains a wealth of opportunities to enhance customer satisfaction.In-depth analysis of customer complaints,and founds that users dissatisfied reasons,is the most importantmeans to enhance customer satisfaction.As the complainant contains large data and text,the traditional manual method requires analysis of each of the text content,so that the process is very cumbersome,the hot spot analysis is not thorough,and the root cause of complaints is difficult to find.In order to improve the analysis efficiency and accuracy,there must be a method for text content analysis.After Investigation and research,at present the customer complaint behavior of customer intelligent analysis system using text classification technology more.In order to ensure the accuracy and efficiency of the project,the project designs much evaluation method for the selection algorithm.Based on the assessment results,the paper chooses LM algorithm which has high accuracy,and introduces anti vibration scheme,to further improves the accuracy and stability of the algorithm.On this basis,through the needs analysis,detailed design,software test and a series of software engineering process,the paper realizes a series of core function,such as intelligent recognition,intelligent analysis and complaints intelligence analysis for potential customer complaints.After going live,the system's accurate rate is up to 70%,which has increased 10%compared with older.Through the system,complaint analysis can get daily or monthly data for intelligent classification.The system improves the complaints follow up time,and reduces the cost of labor compared with artificial monthly analysis before the system.