Research On Improvement And Application Of Quality Function Deployment Model Based On Data Mining

As customer satisfaction continues to occupy an important position in enterprise production,Quality Function Deployment(QFD)as a quality management method aiming at achieving customer satisfaction,are applied by more and more enterprises to manufacture and become a powerful tool to improve product quality and enhance competitive advantage.The key input of QFD is customer demand information.However,due to increasing complexity of customer consumption patterns and product sales channels,the traditional customer demand analysis method slowly shows many problems: on the one hand,the customer demand type is obtained in a single way,mostly for the questionnaire.The survey method,data source and the amount of data are few,so the results are subjective.Besides,the enterprise lacks rationality in the classification and weight distribution of customer needs,they cannot make more targeted product production and targeted plans according to customer needs.In addition,other input information in QFD,such as "association matrix" and "product technical characteristic autocorrelation matrix",have inherent ambiguity and uncertainty.The above factors make the traditional QFD have obvious deficiencies to obtain better application in manufacture.In this background,this paper takes the online customer reviews of e-commerce products as the data source,and focuses on the mining of customer needs,in order to improves the QFD model through the following aspects:(1)First of all,the article uses the consumer comment in the Internet as the data source,and uses the data mining technology as for topic model and association rule mining algorithm to obtain the customer's demand type of the product,thus obtaining more realistic and reliable customer demand information.Through the quantitative Kano model,the customer needs are divided into attractive,basic and one-dimensional attributes,and the final customer demand weights are calculated to find out the importance of different attributes of the products to the customers.(2)After obtaining the customer demand data,the basic importance of the product technical characteristics is determined in the fuzzy environment,and the fuzzy association rule mining algorithm is proposed to analyze the final priority of the technical characteristics.The overall uncertainty of the product characterization and the accuracy of the results is improved.(3)Finally,based on the customer reviews and related technical characteristics of smart phone products in the e-commerce platform,case studies are carried out.By constructing the QFD model,11 customer demand attributes and corresponding 12 products of the smart phone products are finally obtained.Technical characteristics information,and gives priority rankings for smartphone product improvement.The results obtained are more objective,effectively avoiding the ambiguity and uncertainty characteristics of QFD,thus verifying the scientific and practicality of the model.