應用資料探勘技術於智慧型企業資源規劃系統

Abstract

當企業紛紛完成ERP系統的導入應用之後，以其提昇營運效率的競爭優勢便逐漸消失。在知識經濟時代中，要掌握競爭優勢必須先掌握知識。能活化ERP系統蘊藏的有價數位資產，讓資料轉化成資訊、知識與智慧，提昇企業資訊系統的應用境界，將是企業建立差異化競爭優勢的關鍵。因此，本研究提出一套以ERP系統為基礎，並應用資料探勘技術，讓企業邁向商業智慧的系統架構。 本研究整合企業資源規劃、資料倉儲、線上分析處理、資料探勘、人工智慧等五大學域，提出智慧型企業資源規劃系統。本系統架構是以ERP系統所積蓄的既有龐大資料為基礎，經由整合ERP系統與資料倉儲、整合資料倉儲與決策分析、整合決策分析與資料探勘系統等三階段的過程，實際建置涵蓋操作資料、倉儲資料、智慧應用等三個層級的系統架構。資料探勘系統則是應用決策樹演算法與分類預測的模型，以探勘ERP系統中隱含的、有意義的決策資訊。探勘到的知識規則，則以專家系統的法則式知識表現，搭配WEB化方式來展示。 本研究的結果能夠緊密整合ERP系統與資料探勘系統讓ERP系統智慧化。本研究將智慧型企業資源規劃系統，應用於企業製程領域的品質改善，以探勘正確的品質情報知識規則，讓企業了解品質要因的特徵與特性，確實掌握品質問題點。根據實驗分析得知零件短路不良率平均改善率為29.4%、零件空焊不良率平均改善率為35.1%、錫裂脆化不良率平均改善率為33.3%。 本研究所提出的系統架構，可應用於行銷領域的顧客特性分析、目標市場分析、客戶忠誠度分析、預測客戶流失、客戶區隔等方面，協助企業提昇競爭力。

most of the enterprises have already implemented the ERP system , so the competitive advantages using ERP system to enhance managing efficiency have lessened gradually. The enterprises create own competitive advantages as long as they master knowledge in the knowledge economy Age. The key issue that the enterprises create competitive advantages with a difference will depend on whether they take advantage of valuable digit assets in the ERP system , transform data into information , then transform information into knowledge and intelligence , and finally enhance the application level of business information system. This research proposes a system architecture that is based on the application of data mining technology for ERP system and helps enterprises to attain the business intelligence. This research proposes a intelligent ERP system that integrates five theories including enterprise resource planning , data warehouse , online analytical processing , data mining , and artificial intelligence. The system architecture is based on a large number of current datas that were recorded by the ERP system. It consists of three stages including integrated ERP system and data warehouse , integrated data warehouse and decision analysis , integrated decision analysis and data mining system. And the research actually implements a prototyping system consisting of operational data level , warehouse data level , and intelligent application level. The data mining system is actually designed by using the decision tree algorithm and classification model. And it can discover the hidden and meaningful decision information. The results of newfound knowledge rules are represented by the rule-based knowledge representation of expert system and web browser. The result of this research integrates ERP system and data mining system tightly. And it lets ERP system become intelligent management decision system due to the application of discovered knowledge. This research applies the intelligent ERP system to the electron information industry. It is applied to improve the problems of quality and find the correct knowledge rules about quality information. The discovered knowledge rules help enterprises to understand the characterizations of quality , and they let the factors of quality be at the mercy of enterprises. The benefit of intelligent ERP system is to turn this newfound knowledge into actionable results , such as increasing an average improvement rate of 29.4% over the short circuit of electronic parts , increasing an average improvement rate of 35.1% over the parts loosened , and increasing an average improvement rate of 33.3% over the quality of solder. The proposed system architecture can be applied to lots of domains in order to help enterprises to enhance competence , such as customer profiling analysis , targeted marketing analysis , customer loyalty , customer retention , or customer segmentation.