Data Mining And TRIZ Theory Research In Quality Management

Quality is the lifeline of business. In today’s competitive market environment, every enterprise must provide the excellent products or service that complies with market demand to customers. From product design, raw material procurement, manufacturing and processing, product output, the factory distribution, and finally to the hands of customers, each stage will have to go through a strict quality control system, to collect data on all aspects of quality to make a real time processing, and to make quality assessment as well as to improve quality problems continuously. In recent years, customer relationship management has become a business operations concept. It is essential for enterprises to listen to customer needs, and to provide services and to correct corporate defect. The quality is closely related with the customer relationship. It includes maintaining the customer groups, developing new customer base and customer purchasing life cycle and so on. Therefore, enterprises must establish the customer information collecting and rapid response mechanism for customers to reflect the quality problems to make the right, quick response and its treatment.Surface Mount Technology process is a very complex process, which is subject to influenced by personnel, materials, machines, methods, environmental and other factors. Once the problems about product quality are appeared, finding the reason of the quality problems quickly and solving it in time is the key to SMT processes.In this paper, the process from data to decision-making is to be quantified, and the quality control on preliminary result of data mining is to be made. This paper combines the result of analysis with the improved TRIZ theoretical model to establish the diagnosis system for quality problems. Based on the domestic researches, the paper uses the method that combines experiment research with inductive inference to arrange the online quality control and to collect related information in order to find out the reasons affecting the yield of statistical process control of STM. Then the paper uses the data mining decision trees and the method in Six Sigma(DMAIC) and other methods to finds out the optimal parameter of SPC, and finally through the TRIZ theory to resolve the problems in the implementation of decision-making. This paper chooses mobile phone manufacturers as an example to make empirical study for the method about diagnosing and analyzing the quality issues in SMT, and to explain and introduce the overall SMT manufacturing process which includes chemistry SPC and machinery SPC, and to enumerate the common problems on SPC, and to establish the diagnostic analysis method about SMT quality issues. Based on the data mining model, this paper choose the treatment list which is used to improve the action as an example, and generalizes the reason about customer response, and generate a series of problems including generation of “knowledge discovery in databases(KDD)” law and knowledge expression and so on, and finally establish quality issues diagnostics system.