Kalite güvence sistemindeki bazı öğelerin denetim sürecine ilişkin bilgi tabanlı bir uzman sistem yaklaşımı
Başlık çevirisi mevcut değil.
- Tez No: 46446
- Danışmanlar: PROF.DR. AYHAN TORAMAN
- Tez Türü: Doktora
- Konular: Mühendislik Bilimleri, İşletme, Engineering Sciences, Business Administration
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 1995
- Dil: Türkçe
- Üniversite: İstanbul Teknik Üniversitesi
- Enstitü: Fen Bilimleri Enstitüsü
- Ana Bilim Dalı: Belirtilmemiş.
- Bilim Dalı: Belirtilmemiş.
- Sayfa Sayısı: 193
Özet
ÖZET Bu tez çalışmasının amacı; bilgi tabanlı bir uzman sistem yaklaşımı ile ISO 9000 kalite güvence sisteminde yer alan“sözleşmenin gözden geçirilmesi”,“satınalma”ve“taşıma, depolama, paketleme, dağıtım”kalite sistem öğelerinin denetim sürecine ilişkin bir prototip sistem önerisinde bulunmaktır. Önerilen prototip sistem ile küçük ve orta ölçekli sanayi işletmelerinde kalite yönetiminde bilgi tabanlı uzman sistem yaklaşımının uygulamalarına ışık tutulmaya çalışılmıştır. Bu bağlamda çalışma, sekiz bölümden oluşmuştur. Birinci bölümde, kalite yönetiminde bilgi tabanlı uzman sistem yaklaşımının önemi vurgulanmıştır. Yapay zeka ve uzman sistemlere giriş niteliğini taşıyan ikinci bölümde; bu disiplinlere ilişkin temel kavramlar, tarihsel gelişme evreleri, uygulama ve kullanım alanları incelenmiştir. Bu bölümde ek olarak, yapay zeka ve uzman sistemlerin diğer bazı akademik disiplinler ile ilişkilerine de yer verilmiştir. Üçüncü bölümde, bilgi tabanlı uzman sistemlerde bilgi temsil etme yöntemlerinden bir tanesi olan üretim kuralı modeli incelenmiştir. Bu tez çalışmasının uygulama bölümünün kuramsal çatısını oluşturan dördüncü bölümde, uzman sistemlerin ana elemanları ve geliştirme süreci incelenmiştir. Beşinci bölümde, kalite yönetiminde sistem ve uzman sistem yaklaşımı tartışılmıştır. Altıncı bölümde ise, bilgi tabanlı bir uzman sistem yaklaşımı ile kalite güvence sistemindeki bazı öğelerin denetim sürecine ilişkin bir prototip sistem önerisinde bulunulmuştur. Yedinci bölümde, önerilen sistemin orta ölçekli iki sanayi işletmesine uygulanma süreci ve elde edilen sonuçlar sunulmuştur. Sonuçlar ve öneriler bölümünde, bu çalışma çerçevesinde elde edilen yazına ve uygulamaya ait bulgulara yer verilmiştir. Ek olarak, ileriki çalışmalara yönelik önerilerde bulunulmuştur. -IX-
Özet (Çeviri)
SUMMARY A KNOWLEDGE-BASED EXPERT SYSTEM APPROACH FOR THE AUDITING PROCESS OF SOME ELEMENTS IN THE QUALITY ASSURANCE SYSTEM The purpose of this study is to propose a prototype system by means of a knowledge based expert system approach for the auditing process of“contract review”,“purchasing”, and“handling, storage, packaging, delivery”quality system elements in the ISO 9000 quality assurance system. In regard to the proposed system, it is dealt with the application of knowledge based expert system approach of quality management in small and medium sized manufacturing organizations. In this connection, this study consists of eight chapters. In the first chapter, the importance of knowledge based expert systems in quality management is emphasized. In the second chapter, which is the introduction of artificial intelligence and expert systems, the basic concepts, historical development stages, application, and usage areas of these diciplines are examined. In addition, the relationship of th ; artificial intelligence, expert systems, and some other academic diciplines are explained in the same chapter. One of the knowledge representation paradigms of expert system, called the production rule method, is examined in the third chapter. In the fourth chapter, the theoretical framework of the application of this study is explained, and the main elements and the development process of expert systems are studied. Furthermore, expert system software family is examined. In the fifth chapter, the system and the expert system approach in quality management is discussed. In the sixth chapter, a prototype system for the auditing of some quality elements of the quality assurance system is proposed. In the seventh chapter, the application of the proposed system in two of the medium sized manufacturing companies is presented and the consultation processes and findings are presented as well. In the final chapter, where the results and proposals are presented, the findings from the literature review and the application of this study are introduced. Furthermore, the proposals for the future work are discussed. Computers and other types of information systems have been used to support decision making in management science for over four decades. Electronic Data Processing Systems (EDPSs) appeared in the mid-1950s. ManagementInformation Systems (MISs) followed this evolution in the 1960s. Office Automation Systems (OASs) were developed mainly in the 1970s. Perhaps the most attractive computer based information system in the late 1980s and 1990s was Expert Systems (ESs). Expert systems are one of the subarea of applied artificial intelligence. Artificial intelligence is the field of computer science concerned with designing intelligent computer systems; that is the science of making machines do things that would require intelligence in cases done by human beings. According to these definitions, artificial intelligence focuses on two basic ideas. These ideas are stuyding the thinking processes of humans and representing these processes by means of machines like computers, robots, etc. When researchers realized the fact that the knowledge and the manipulation of this knowledge provided a reasonable means to the general problem solving mechanism, expert system was introduced. This conceptual change openly showed that the specific knowledge about the problem domain has considerable importance on the programming area. According to Ignizio (Ignizio, 1991), an expert system is a model and associated procedure that exhibits within a specific domain, a degree of expertise in problem solving which is comparable to that of a human expert. Within this definition, expert system model is the representatian of human knowledge in a knowledge base. In this study, this expert system concept is justified. Although no certain date is known for the origin of artificial intelligence, the year 1956 is accepted as a very important date for the historical evolution of artificial intelligence. In 1956,“artificial intelligence”term was suggested by John McCarthy, who was one of the organizers of Darmouth Conference conducted by Darmouth College in New Hampshire - Hanover, U.S.A. In 1957, McCarty developed the first artificial intelligence programming language called LISP (List Processing Programming). During the 1960s, computer scientists developed a number of general problem solving mechanisms by applying these mechanisms to“real life”problems resulted in dismal failure. In those days, researchers realized that the solution of a problem was highly influenced by the specific knowledge that was in the context of the problem. Within this conceptual revolution, the first expert system called DENDRAL was developed in the late 1960s and early 1970s at Stanford University in U.S.A. In 1970s, PROLOG (Programming Logic) which has been widely used in Europe and Japan was developed by Alan Colmerauer. In 1970s, the general problem solving mechanisms shifted towards the inference mechanisms in the specific type of problem areas. Today's modern expert systems are based on this approach. Since 1980s, expert systems have been used widely in the commercial areas. Expert systems might be considered for any situation where it is valuable to preserve expertise. Because of this reason, expert systems can provide major benefits to users. Some of the benefits of expert systems are the following: -XI-- Cost reduction - Increased output and productivity - Increased quality - Flexibility - Reliability - Conceptuality - Capturing scarce expertise - Working with incomplete and uncertain information - Educational benefits - Timeliness in decision making Although the usage of expert systems provide considerable benefits to the users, they should not be considered as a magic. There are lots of limitations in the expert system area. Knowledge is not always readily available and expertise is hard to extract from humans. However, they work well in some cases in very narrow domains. On the other hand, production systems developed by A. Newell and H.C. Simon are modular knowledge representation schemes in expert systems modelling the problem solving process of a human expert by means of knowledge representation and search algorithms. The basic idea of these systems is that knowledge is represented as rules, called productions. The rules are the form of condition - action pairs, such as“IF”a condition occurs,“THEN”an action occurs. Production systems contain three main elements. These are production rule sets, working memory, and inference engine. The inference process can be driven for solving a problem using several control strategies. But, the two main strategies called forward chaining in the forward mode, and backward chaining in the backward mode, are the most widely used search control strategies. An expert system is generally consisted of six main elements. They are knowledge acquisition, knowledge base, inference engine, working memory, user interface, and explanation facility. The knowledge base contains facts and rules that embodies the expert knowledge. Inference engine decides the order and the organization in which inferences are made. Working memory contains facts that emerge from consultation with the knowledge base. This consultation is driven by the inference engine containing the inference strategies and controls which a human expert uses during the problem solving processes. The user is interfaced within the system via menus, graphs or natural language processing. When a user interfaces with the system, within the consultation process, the user respones are inferred by the inference engine driving the rules and facts from the knowledge base. On the other hand, developing of an expert system may differ from the point of views of the software used, the size of the problem domain and the resources allocated to this project. Generally, this process can be developed by one of the three methods which are custom development, semi-custom development, or expert system application packages. In the custom development, the system is built from scratch using one of the artificial intelligence development languages. The semi-custom development method is based on the modelling of the knowledge base using an expert system shell. In the third method, the ready made expert -XII-system packages are used. But, some adjustment may be made to fit the exact application needs. The basic phases in the development process of an expert system are the following: - Definition of the problem domain, - Selection of expert(s), - Conceptual design and feasibility study, - Selection of software and hardware environment, - Knowledge acquisition, knowledge representation, and inference engine, - Building a prototype, - Performance evaluation, validation, and verification, - Acceptance by the user, - Implementing expert system, - Documentation, maintenance, and security. In this study, the above mentioned development processes are implemented in the three quality system requirements of ISO 9000 quality assurance system. Although expert system approach have been widely utilized in medicine, geology and military, realm of interest is begining to expand in management and quality management as well. Since quality management is a vital and growing concern of bussiness in the global market place, expert systems have recenty become an important part of this evolution, and will be more commonly used in this area in the future. According to this point of view, by means of using an expert system approach, a prototype is proposed for the auditing process of“contract review”,“purchasing”, and“handling, storage, packaging and delivery”quality assurance system requirements of ISO 9000. It is obvious that most of the small and medium sized companies should implement the quality assurance system in their organizations, because of the forces of large sized companies. In the globalization processes, the existence of“zero defect production”is unavoidable to tackle with high competition. Therefore, this type of production can be made by means of a system approach in quality management. By means of using an expert system approach, it is possible to preserve the experts heuristics, rules, and facts in a knowledge base to present the users, who are not experts or in the training stages. In Turkey, there are more than two hundred and fifty companies which have achieved ISO 9000 certification. Furthermore, there are some companies that provide certification and consultation for ISO 9000. Therefore, there are human experts who capture their expertise in order to use this expertise for the companies. By means of using the proposed expert system approach, it is aimed to present expertise related to the mentioned three quality system elements to the users. This approach called ESAQAS (An Expert System Approach in Quality Assurance System) will only guide the user during the preperation stages of the three quality system elements of ISO 9000 quality assurance system. It neither encompasses the whole processes of the quality assurance system nor provides the certification. But, it is obvious that this modular approach will enlighten the completion of the whole system within a group work. ESAQAS has five main elements called knowledge acquisition, knowledge base, inference engine, working memory, and user interface. -xiii-During the development processes of the proposed expert system approach, the theoretical framework taken place in the literature is followed. The first phase which is the definition of problem domain is divided in mainly two parts which are researchers' evaluation and experts' evaluation. These evaluation processes are accomplished from the point of view of availability of human experts, symbolic reasoning and the determination of the limits of problem domain being considered. At the end of these processes,“contract review”,“purchasing”, and“handling, storage, packaging, and delivery”quality system requirements are found to be the most proper elements to capture expertise, and to represent knowledge in the form of heuristic, and rules, since the whole system requires more resources like financing, timeliness, and etc. On the other hand, the quality system auditing process requires visual evidence fully. However, these three elements may require less visual evidence than that of others. After defining the problem domain, the selection of expert(s) phase is accomplished. This phase also split into two parts, which are selecting the written resources and human experts. Although expert system in quality management is in the begining stages, there are a few written resources, related to this topic. On the other hand, the quality managers in the eight firms in Istanbul that have received ISO 9000 certification have been chosen for the knowledge acquisition of the proposed approach. In addition to this, the experts from the certification and consultation companies were interviewed for this phase. In the conceptual desing and feasibility study phase, the operational feasibility like the purpose and the architecture of the system is performed. Technical feasibility like hardware and software availability, knowledge and data availability and knowledge representation paradigm are defined. Finally, the inferencing control strategy, and the framework of the proposed system are defined at the same stage. For the development environment, LEVEL5 expert system shell was used as a software, because of its availability. On the other hand, an IBM compatible PC with a minimum 51 2K RAM on MS-DOS operating system was used since LEVEL5 has been designed to run on this type of system. The fifth phase, which are knowledge acquisition, knowledge representation, and inference engine took a very long time. The knowledge acquisition processes are realised in eight phases. These are the following: 1) Preliminary interviews and questionnaires, 2) Examining the preliminary knowledge acquired, 3) The secondary interviews, 4) Editing the knowledge acquired, 5) Evaluation of the literature, 6) Rapid protoyping, 7) Questionnaires and rating scales, 8) Verification of the knowledge by the experts and conflict resolution. In building a prototype phase, acquired knowledge is encoded into the four knowledge bases called DENETIM.PRL, DENETİMİ. PRL, DENETIM2.PRL and DENETIM3. PRL by means of“Production Rule Language”of LEVEL5. Both backward and forward chaining control staretgies were used in this knowledge bases. DENETIM.PRL knowledge base contains the knowledge related to the -xiv-identification and the beginning process of the consultation. On the other hand, DENETİMİ. PRL, DENETIM2.PRL and DENETIM3. PRL knowledge bases contain the knowledge in the order of“contract review”,“purchasing”, and“handling, storage, packaging and delivery”audit consultation. Each knowledge base can run after the compilation process. During the consultation session, the user is asked several questions to determine the current position of preperation stage of each quality system requirement considered. At the end of the consultation session, the user is informed about the current possition of the system. In adition to this, the system is inferred to the total value and the total score of the auditing process being considered. In the verification and the validation phases, various techniques are used. Verification was performed according to the criteria of knowledge acquisition, formal specification, and tool facilities. On the other hand, the validation was performed by the criteria which are project experts, other experts and actual field usage. Both the verification and the validation phases were done during the development and application processes of the proposed approach. Finally, the proposed prototype system is applied in two of the medium sized manufacturing companies, which are the suppliers of the leading automotive sectors mainly. The companies are Kormetal Sanayi ve Ticaret A.Ş. and Net Civata ve Vida Sanayi A.Ş. Kormetal Sanayi ve Ticaret A.Ş. produces light alloy wheels for cars. On the other hand, Net Civata ve Vida Sanayi A.Ş. produces fastening components for the automative, electrotechnical, and mechanical industries. Both of the firms are in the stage of preparation for ISO 9002 certification. After presenting ESAQAS approach, the system is found to be reliable, consistent and timely, and may be cheap for the three mentioned quality system elements by the experts who are in charge of ISO 9000 preperation. The previous literature on quality management lacked the study of knowledge acquisition, knowledge representation, and inference engine. In addition to these; the verification, the validation, and the accetance of users phases have been also not dealt with in the literature. In the proposed system, knowledge acquisition was performed by both the traditional knowledge acquisition techniques and the techniques proposed by the reseracher. Applying these techniques to the knowledge acquisition process took a long time and effort. On the other hand, knowledge representation and inferencing mechanisms are broadly studied. The evaluation which is based on some verification and validation criteria showed that the system is reliable, consistent, and timely for the preperation stage of quality assurance system. For future work, the system can be enriched with more knowledge for the whole system and for the December 1994 version as modular. Furthermore, the sectoral editing can be done with the explanation facility. By means of this, the future work can be used both for the preparation and education purposes. As a consequence, the proposed protoype system has the guidance attributes to the small and medium sized manufacturing companies in the preperation stages. -xv-
Benzer Tezler
- Kalite, kalite güvence ve tekstilde uygulaması
Quality, quality assurence and application on textile
İRFAN GÖNÜLTAŞ
Yüksek Lisans
Türkçe
2008
İşletmeKadir Has Üniversitesiİşletme Ana Bilim Dalı
YRD. DOÇ. DR. DİNA ÇAKMUR YILDIRTAN
- Kalite güvence sistemi ISO 9001 uygulaması
Quality assurance systems and ISO 9001 practise
MUSTAFA GÜMÜŞ
Yüksek Lisans
Türkçe
2002
Endüstri ve Endüstri MühendisliğiSakarya ÜniversitesiEndüstri Mühendisliği Ana Bilim Dalı
DOÇ. DR. M. SABİH AKSOY
DOÇ. DR. ORHAN TORKUL
- Quality assurance policies in higher education in Turkey: The perceptions of the key stakeholders
Türkiye yükseköğretiminde kalite güvencesi politikaları: Önemli paydaşların görüşleri
GÜLŞAH KISABACAK BAŞGÜRBOĞA
Doktora
İngilizce
2022
Eğitim ve ÖğretimBoğaziçi ÜniversitesiEğitim Bilimleri Ana Bilim Dalı
PROF. DR. FATMA NEVRA SEGGIE
- Educational reforms in Ethiopia: From the imperial era to the present
Etiyopya'da eğitim reformları: Emperyal dönemden günümüze
SALİH AHMED MAHAMMODA
Doktora
İngilizce
2022
Eğitim ve ÖğretimNecmettin Erbakan ÜniversitesiEğitim Bilimleri Ana Bilim Dalı
PROF. DR. İSA KORKMAZ
- Yeni gemi inşaatında kalite güvence modeli olarak ISO 9001
Başlık çevirisi yok
SEMA GÜÇER
Yüksek Lisans
Türkçe
1998
Gemi Mühendisliğiİstanbul Teknik ÜniversitesiGemi İnşaat Mühendisliği Ana Bilim Dalı
DOÇ. DR. OSMAN AZMİ ÖZSOYSAL