企業導入智慧型資訊系統效益分析之研究

Abstract

隨著時代的進步與資訊技術的發達，越來越多的企業大量的採用資訊系統來輔助其日常的運作，從個人電腦裡的應用程式到龐大企業體內的ERP，不論是資料的處理或是生產設備的控制都需要資訊系統的幫助。在知識經濟的時代選擇合適的資訊系統收集資訊，找出有利用價值的資訊，並且有效的使用人工智慧技術於資訊系統中，以利於決策分析並增加企業的競爭力是值得深入探討的問題。 資訊系統幫助企業日常運作，而智慧型資訊系統使用人工智慧技術來增進資訊系統運作效能並有效提昇企業機能。本論文以本體論的技術用於使用者需求分析，定義本體論中的類別、屬性及實例，依此本體論來設計智慧型資訊系統之架構。並依使用者需求建立一個符合企業需求之資料倉儲系統，以整合企業內部異質資料的來源，提供使用者一個Web-based的使用介面。此資料倉儲系統將查詢及報表標準化，提供整合性的報表，並可作為資料探勘的資料。另使用資料探勘技術於製造流程中，利用資料探勘技術對製造資料進行演算與推論，可以得知製造資料變數之間的關係。 本研究結果提出一個智慧型資訊系統的架構，其使用本體論技術讓使用者需求分析更能明確的表達，系統功能更符合使用者的實際需求，在系統需要升級及使用者需求改變時，系統的更新更容易且迅速。運用資料倉儲系統能有效降低產品分析週期，提供完整的產品分析資料，使得產品的分析資訊精確且迅速的完成，可作為主管做決策分析的輔助資訊。並使用資料探勘技術在製造流程中，結合生產資料分析與參數設定，可提升產品良率及縮短製造週期，有效的改善生產效能增加企業競爭力。

In modern society, more and more enterprises use information systems, from applications in personal computers to Enterprise Resource Planning in whole enterprises to aid their daily operations; whether the task is data processing or manufacture equipment controlling, it needs the assistance of information system. In the knowledge economy era, using suitable information systems to collect data and discover valuable information and integrate artificial intelligence with information systems for decision supports is an important issue worthy of being discussed deeply. Information aid daily operations of enterprises; intelligent information system uses artificial intelligence to enhance system performance and improve the advantage of enterprise competition. In this research, we use ontology for user requirements analysis; define the class, slot and instance of the ontology; design the system architecture depending on the ontology. We build a data warehouse system according to enterprises’ requirements which are integrated with different data sources in the enterprise and support a web-based interface. It would also support the standard queries and reports, summary tables, and data set of data mining. We apply data mining technology to manufacturing process. It allows us to make predictions in the presence of relationships among a set of variables in manufacturing process. In this research, we propose an intelligent information system architecture, which uses an ontology-based user requirements method to express user requirements clearly and definitely. That would make functions of information system closer to real need. The data warehouse system would reduce the cycle time and support the complete analysis information of production. It makes analysis of production to process quickly and correctly. Moreover, the analysis information would help managers to make decisions. We use data mining technology to improve the efficiency of manufacturing processes, in which combine the data mining result of history data and manufacturing parameters setting. It can improve the yield rate of the manufacturing to improve the advantages of enterprise competition.