• 沒有找到結果。

問題解決之知識支援

N/A
N/A
Protected

Academic year: 2021

Share "問題解決之知識支援"

Copied!
65
0
0

加載中.... (立即查看全文)

全文

(1)

國 立 交 通 大 學

資訊管理研究所

博士論文

問題解決之知識支援

Knowledge Support for Problem-solving

研 究 生: 柯 志 坤

指導教授: 劉 敦 仁 博士

(2)

國 立 交 通 大 學

資訊管理研究所

博 士 論 文

問題解決之知識支援

Knowledge Support for Problem-solving

研 究 生:柯志坤

研究指導委員會:王朝煌 博士

李瑞庭 博士

羅濟群 博士

楊 千 博士

指導教授:劉敦仁 博士

中 華 民 國 九 十 五 年 十 月

(3)

問題解決之知識支援

Knowledge Support for Problem-solving

研 究 生:柯志坤

Student:Chih-Kun Ke

指導教授:劉敦仁

Advisor:Dr. Duen-Ren Liu

國 立 交 通 大 學

資 訊 管 理 研 究 所

博 士 論 文

A Dissertation

Submitted to Institute of Information Management College of Management

National Chiao Tung University in Partial Fulfillment of the Requirements

for the Degree of

Doctor of Philosophy in Information Management October 2006

Hsinchu, Taiwan, the Republic of China

(4)

Knowledge Support for Problem-solving

Student: Chih-Kun Ke Advisor: Dr. Durn-Ren Liu Institute of Information Management,

National Chiao Tung University

Abstract

Problem-solving is an important process that enables corporations to create competitive business advantages. Traditionally, Case-Based Reasoning (CBR) techniques have been widely used to help workers solve problems. However, conventional approaches focus on identifying similar problems without exploring the information needs of workers and relevant context of situation during the problem-solving process. Such processes are usually knowledge intensive tasks; therefore, workers need effective knowledge support that gives them the information necessary to identify the causes of a problem and enables them to take appropriate action to resolve the situation. In this work, we propose a mining-based knowledge support system for problem-solving. Based on CBR and data mining techniques, in addition to adopting CBR techniques to identify similar situations and the action taken to solve them, the proposed system employs text mining (Automatic Indexing) techniques to extract the key concepts of situations and actions. These concepts form profiles that model workers’ information needs when han-dling problems. Effective knowledge support can thus be facilitated by providing workers with situation/action-relevant information based on the profiles. Moreover, association rule mining is used to discover hidden knowledge patterns from historical problem-solving logs. The dis-covered patterns identify frequent associations between situations and actions, and can there-fore provide decision-making knowledge, i.e., appropriate actions for handling specific tions. We develop a prototype system to demonstrate the effectiveness of providing situa-tion/action relevant information and decision-making knowledge to help workers solve prob-lems. Furthermore, based on CBR, data mining, and rule inference techniques, the con-text-based situation identified by CBR techniques provides effective concon-text-based knowledge documents according to the context-based profile. The hidden knowledge patterns are discov-ered to identify inferred associations between situation features and actions, and can therefore provide context-based relevant knowledge. A prototype system is developed to demonstrate the effectiveness of providing inferred knowledge.

(5)

Acknowledgement

誠心地感謝口試委員王朝煌教授、李瑞庭教授、羅濟群教授以及楊千教授,謝謝您 們在百忙之中,於論文口試期間,仍細審論文並給予指正與建議,使得本論文能夠更臻 至完善,在此深表感激。另外,感謝論文指導老師劉敦仁教授,這六年多的指導與照顧。 花了六年多完成了這博士學程,現實的壓力讓我感覺不到任何的驕傲與光榮,不知 道拿到畢業證書的那一刻,眼淚會不會從眼角流下。現在的心境充滿的是感謝還是惆 悵,實在很難分的清楚,只希望往後的日子裡,社會的經歷能告訴我在交大受的訓練是 值得感謝的。另外,想要感謝的親朋好友們,你們名字已經烙印在我的心中,不用在這 邊強調,就讓我們將路走的更長遠吧。 志坤 2006/10

(6)

Contents

Abstract ...i

Acknowledgement...ii

Contents ...iii

List of Figures ... v

List of Tables ...vi

Chapter 1. Introduction ... 1

1.1. Motivation ...1

1.2. Goals...2

1.3. Contributions ...2

1.4. Organization ...4

Chapter 2. Related Work ... 5

2.1. Knowledge management and knowledge retrieval...5

2.2. Problem solving and case-based reasoning ...6

2.3. Information retrieval in a vector space model ...7

2.4. Data mining ...8

2.5. Context-awareness...8

2.6. Rule inference with certainty factor ...9

Chapter 3. Knowledge Requirements of Knowledge Support for Problem-solving... 10

3.1. The problem-solving process ...10

3.2. Knowledge requirements for problem-solving... 11

3.3. Context-based knowledge requirements for problem-solving ...12

Chapter 4. Knowledge Support based on Case-based Reasoning and Data Mining.. 14

4.1. Knowledge support framework for problem-solving ...14

4.2. Discovery of problem-solving knowledge ...16

4.2.1. Data preprocessing for knowledge discovery...17

4.2.2. Situation/action identification and case-based reasoning...18

4.2.3. Discovery of situation/action profiles ...21

4.2.4. Discovery of knowledge patterns ...21

4.3. Knowledge support for problem-solving...23

4.3.1. Knowledge support network...23

4.3.2. Knowledge recommendation...26

(7)

Chapter 5. Knowledge Support based on Case-based Reasoning, Data Mining, and

Rule Inference ... 30

5.1. Context-based knowledge support framework for problem-solving...30

5.2. Discovery of context-based problem-solving knowledge ...32

5.2.1. Context-based situation identification and case-based reasoning ...32

5.2.2. Discovery of context-based situation profiles ...40

5.2.3. Discovery of context-based knowledge patterns...40

5.2.4. Knowledge recommendation...44

5.3. System implementation ...46

5.4. Discussions and comparisons ...48

5.4.1. Discussions ...48

5.4.2. Comparisons ...49

Chapter 6. Conclusions and Future works ... 52

6.1. Summary...52

6.2. Future works...53

(8)

List of Figures

Fig. 1: A problem-solving process for a production line ... 12

Fig. 2: Knowledge support framework for problem-solving... 14

Fig. 3: The procedures of knowledge discovery process... 17

Fig. 4: The procedures of knowledge recommendation ... 24

Fig. 5: Decision-making knowledge patterns in a knowledge support network ... 25

Fig. 6: Dependency knowledge pattern in a KSN ... 25

Fig. 7: Situation profile and relevant documents... 27

Fig. 8: A generic water supply problem-solving process ... 28

Fig. 9: Decision-making knowledge patterns for the water supply problem ... 29

Fig. 10: The adapted system framework of context-based knowledge support ... 30

Fig. 11: An example of inference process ... 38

Fig. 12: An example of context-based decision-making knowledge pattern... 41

Fig. 13: An example of context-based dependency knowledge pattern ... 43

Fig. 14: Relevant context-based situation profile of situation EI_S4... 44

Fig. 15: Context-based situation profiles and relevant documents of the pattern ... 45

Fig. 16: Context-based knowledge patterns in a chain reaction across different stages ... 46

Fig. 17: A prototype system of context-based problem-solving knowledge support ... 47

(9)

List of Tables

(10)

Chapter 1. Introduction

1.1. Motivation

Problem-solving is an important process that enables corporations to create competitive advantages, especially in the manufacturing industry. Case-based reasoning (CBR) tech-niques (Chang et al., 1996; Kohno et al., 1997; Park et al., 1998; Yang et al. 2004) have been widely used to help workers solve problems. For example, based on these techniques, a decision support system was developed to facilitate problem-solving in a complex produc-tion process (Park et al., 1998). CBR techniques have also been used to implement a self-improvement helpdesk service system (Chang et al., 1996), and integrated with the ART-Kohonen Neural Network (ART-KNN) to enhance fault diagnosis in electric motors (Yang et al., 2004).

Conventional CBR approaches focus on identifying similar problems without explor-ing the information needs of workers and relevant context of situations durexplor-ing prob-lem-solving tasks. Probprob-lem-solving is a complex process that includes a series of uncertain situations and operational actions. Moreover, it is usually knowledge intensive and workers need to access relevant information in order to identify the causes of a situation and take appropriate action to solve it. Situation features are usually occurred according to the context characteristics of problem. Due to the uncertain features of situations, several causes and possible solutions may exist for a specific situation. For example, in a production process, a significant decline in performance may be due to poor materials, inexperienced workers, or faulty machinery. Thus, possible solutions would include replacing the poor materials, re-training the workers, or repairing the faulty machinery. The causes and possible solutions are usually hidden in relevant data resources and difficult to extract. In such uncertain envi-ronments, situation features collected by system are usually partial or incomplete. Workers need to use knowledge gathered and inferred from relevant context information and previous problem-solving experience to clarify the causes and take appropriate action effectively. Thus, identifying similar cases through CBR is not sufficient to solve problems. An effective knowledge support system is essential so that workers have the information necessary to identify the causes of a problem and take appropriate action to solve it.

(11)

1.2. Goals

According to the motivation, this work lists major goals as follows: z Analyze collected attributes of situation/action of problem-solving;

z Based on CBR and data mining techniques, design a system framework of knowledge support for problem-solving;

z Identify similar situations/actions by CBR;

z Discovery of situation/action profile and knowledge patterns;

z Construct a knowledge support network for knowledge recommendation;

z Implement a prototype system to demonstrate the effectiveness of proposed framework;

z Analyze collected attributes and features of situation for problem-solving;

z Based on CBR, data mining, and rule inference techniques, design a system framework of context-based knowledge support for problem-solving;

z Identify similar context-based situations by CBR;

z Discovery of context-based situation profile and relevant knowledge (e.g., knowledge patterns and relevant knowledge documents);

z Implement a prototype system to demonstrate the effectiveness of proposed framework.

1.3. Contributions

In this work, we propose a mining-based knowledge support system for problem solving. Besides adopting CBR to identify similar situations and the action taken to solve them, we adopt text mining (Automatic Indexing) and rule inference techniques to com-pensate for the shortcomings of CBR technique. For specific situations or actions, their situation/action attributes, features, context characteristics and relevant information (documents) accessed by workers is recorded in a problem-solving log. Historical codified knowledge (textual documents), i.e., experience and know-how extracted from previous problem-solving logs, can provide valuable knowledge for solving the current problem.

The proposed system employs Information Retrieval (Automatic Indexing) techniques to extract the key concepts of relevant information necessary to handle a specific situation or action. The extracted key concepts form a situation/action profile that models the informa-tion needs of workers for a specific problem-solving task. The system can then uses the

(12)

situation/action profile to gather existing and new relevant knowledge documents for

spe-cific situation/action. We employ association rule mining methods to discover deci-sion-making knowledge rules about frequently adopted actions taken to handle specific situations. These rules are generated as knowledge support to help workers take the appro-priate action to solve a specific situation. Furthermore, the problem-solving process in-cludes a series of uncertain situations and operational actions, and preceding situations or actions may trigger subsequent problem situations. Therefore, workers need to gather such triggering information (chain reactions) to determine appropriate action. For example, if an unstable system causes production to decline, the solution may be to reboot the system. However, this may result in breakage of materials, which would increase production costs. The proposed approach applies sequential pattern mining methods to discover dependency knowledge which represents frequent chain-reactions. The knowledge helps workers make appropriate action plans. The discovered profiles and knowledge rules are used to construct a knowledge support network, which provides workers with relevant situation/action in-formation, as well as decision-making and dependency knowledge. A prototype system is developed to demonstrate the effectiveness of the knowledge support network.

Moreover, we adapt system framework to provide context-based knowledge support for problem-solving. The adapted system employs constraint-based association rule mining methods to discover context-based inference rules from the problem-solving log. Con-text-based inference rules identify inferred associations between situation features and relevant context characteristics. Based on the discovered context-based inference rules, the system infers more situation features to assist CBR in situation identification. The proposed system employs Information Retrieval (Automatic Indexing) techniques to extract the key concepts of relevant information necessary to handle a specific situation. The extracted key concepts form a context-based situation profile that models the information needs of work-ers for handling problem situation in certain context. The system can then uses the

con-text-based situation profile to gather existing and new relevant knowledge documents for

specific situation according to the context information. Furthermore, the adapted system continually infers situation features to form the context-based knowledge patterns which provide workers with relevant inferred knowledge (inferred situation features and relevant context-based inference rules), as well as context-based decision-making and dependency knowledge.

(13)

1.4. Organization

The remainder of this work is organized as follows. Chapter 2 reviews related works on knowledge discovery and problem-solving. Chapter 3 introduces the knowledge require-ments of knowledge support for problem-solving. Chapter 4 describes the knowledge sup-port based on CBR and data mining techniques, including knowledge supsup-port framework for problem-solving, discovery of problem-solving knowledge, knowledge support for prob-lem-solving, and a prototype system implementation. Chapter 5 illustrates the knowledge support based on CBR, data mining, and rule inference techniques, including context-based knowledge support framework for problem-solving, discovery of context-based prob-lem-solving knowledge, the prototype system, discussions, and comparisons. Finally, we summarize this work and describe the future works in Chapter 6.

(14)

Chapter 2. Related Work

The related literature covers knowledge management, problem-solving, case-based reasoning, information retrieval, data mining techniques, and context-awareness.

2.1. Knowledge management and knowledge retrieval

AI techniques have advanced knowledge management, including knowledge acquisi-tion, knowledge repositories, knowledge discovery, and knowledge distribution (Liebowitz, 2001). Knowledge acquisition captures tacit and explicit knowledge from domain experts (Kohno et al, 1997; Klemettinen et al., 1997), while knowledge repositories formalize the outcomes of knowledge acquisition and integrate knowledge in distributed corporate envi-ronments (Georgalas, 1999). Taxonomy and mapping mechanisms are used to represent relevant knowledge and construct a framework for building a knowledge repository (Chakrabarti et al., 1997). Knowledge discovery and mining approaches explore relation-ships and trends in the knowledge repositories to create new knowledge. In addition, heu-ristic mechanisms, such as proactive knowledge delivery and context-aware knowledge retrieval, are used to enhance knowledge distribution (Abecker et al., 2000).

A repository of structured, explicit knowledge, especially in document form, is a codified strategy for managing knowledge (Davenport & Prusak, 1998; Gray, 2001). However, with the growing amount of information in organization memories, knowledge management systems (KMS) face the challenge of helping users find pertinent information. Accordingly, knowledge retrieval is considered a core component in accessing information in knowledge repositories (Kwan & Balasubramanian, 2003; Fenstermacher, 2002). Translating users’ information needs into queries is not easy. Most systems use Information Retrieval (IR) techniques to access organizational codified knowledge. The use of Infor-mation Filtering (IF) with a profiling method to model users’ inforInfor-mation needs is an ef-fective approach that proactively delivers relevant information to users. The technique has been widely used in the areas of Information Retrieval and Recommender Systems (Her-locker & Konstan, 2001; Middleton et al., 2004; Pazzani & Billsus, 1997). The profiling approach has also been adopted by some KMS’ to enhance knowledge retrieval (Abecker et al., 2000; Agostini et al., 2003; Davies et al., 2003), whereby information is delivered to task-based business environments to support proactive delivery of task-relevant knowledge

(15)

2.2. Problem solving and case-based reasoning

Problem solving is the thought process that resolves various difficulties and obstacles spread in the gap between the current problem and its desired solution. (Heh, 1999). Past experience or knowledge, routine problem-solving procedures, and previous decisions can be used to enhance problem-solving. Liao (2002) investigates the types of knowledge used for problem-solving and suggests the circulation of knowledge to avoid knowledge inertia. Although a knowledge-based architecture that incorporates case-based, rule-based, and heuristic-based approaches is proposed for managing problem-solving knowledge and dealing with knowledge inertia, the details of the system are not presented.

Various approaches that integrate AI techniques have been proposed to support prob-lem solving. Case-based reasoning (CBR), which has been widely used to help workers solve problems, is the process of solving a given problem based on the knowledge gained from solving previous similar problems (Allen et al., 2002). Most CBR systems include the following steps: case representation and storage, precedent matching and retrieval, adapta-tion of the retrieved soluadapta-tion, validaadapta-tion of the soluadapta-tion, and case-base updating to include the information gained from solving the new problem. The CBR approach was used to im-plement a self-improvement helpdesk service system (Chang et al., 1996), and a CBR-based decision support system was developed for problem-solving in a complex production process (Park et al., 1998). More recently, Yang et al (2004) proposed integrating the CBR approach with ART-Kohonen neural networks (ART-KNN) to enhance fault diagnosis in electric motors. Moreover, RBCShell was introduced as a tool for constructing knowl-edge-based systems with CBR (Guardati, 1998), whereby previously solved problems are stored in the case memory to support problem-solving in new cases.

Existing studies focus on using case-based reasoning to identify similar previous cases and derive a solution for a new case from previous problem solutions. In a complex pro-duction process, problem-solving is usually knowledge intensive and requires effective knowledge support to provide workers with the necessary information to identify the causes of situations and taking appropriate action to solve them. However, identifying similar cases among previous problem cases is not sufficient to satisfy workers’ information needs for solving a new problem. The required knowledge is usually hidden in various codified knowledge documents that must be proactively delivered to workers. The CBR approach does not provide such problem-relevant documents for knowledge-intensive problem solv-ing.

(16)

2.3. Information retrieval in a vector space model

The key contents of a codified knowledge item (document) can be represented as a term vector (i.e., a feature vector of weighted terms) in n-dimensional space, using a term weighting approach that considers the term frequency, inverse document frequency, and normalization factors (Salton et al., 1988). The term transformation steps, including case folding, stemming, and stop word removal, are performed during text pre-processing (Salton et al., 1971; Poter, 1980; Witten et al., 1999). Then, term weighting is applied to extract the most discriminating terms (Baeza-Yates et al., 1999). Let d be a codified knowledge item (document), and let dr= <w(k1, d), w(k2, d), …, w(kn, d)> be the term vector of d, where w(ki,

d) is the weight of a term ki that occurs in d. Note that the weight of a term represents its degree of importance in representing the document (codified knowledge). The well-known

tf-idf approach, which is often used for term (keyword) weighting (Poter, 1980), assumes that

terms with higher frequency in a document and lower frequency in other documents are better discriminators for representing the document. Let the term frequency be the occurrence frequency of term k

) , ( dk tf i

i in d, and let the document frequency represent the number of documents that contain k

) (ki df

i. The importance of ki is proportional to the term fre-quency and inversely proportional to the document frefre-quency, which is expressed as Equa-tion 1:

(

( , ) log( ( ) 1)

)

( , ) (log ( ) 1) 1 ) , ( 2 × + + × =

i i i i i i k df N d k tf k df N d k tf d k w r , (1)

where N is the total the number of documents. Note that the denominator on the right-hand side of the equation is a normalization factor that normalizes the weight of a term.

Similarity measure: The cosine formula is widely used to measure the degree of similarity

between two items, x and y, by computing the cosine of the angle between their corre-sponding term vectors and x yr , which is given by Equation 2. The degree of similarity is

higher if the cosine similarity is close to 1.

y x y x y x y x sim( , )=cosine(r, r)= rrrr (2)

(17)

2.4. Data mining

Data mining, which has become an increasingly important research area, involves several tasks, including association rule mining, sequential pattern mining, clustering, clas-sification, and prediction (Chen et al.,1996; Han & Kamber, 2000). We adopt association rule mining and sequential pattern mining to extract knowledge patterns from previous problem-solving instances.

Association rules mining. Association rule mining tries to find an association between two

sets of products in a transaction database. Agrawal et al. (1993) formalized the problem of finding association rules as follows. Let I be a set of product items and D be a set of trans-actions, each of which includes a set of products that are purchased together. An association rule is an implication of the formXY, whereXI,Y ⊂ , and I X ∩Y. X is the an-tecedent (body) and Y is the consequent (head) of the rule. Two measures, support and confidence, are used to indicate the quality of an association rule. The support of a rule is the percentage of transactions that contain both X and Y, whereas the confidence is the fraction of transactions containing X that also contain Y.

Sequential pattern mining. The input data is a set of sequences, called data-sequences. A

data-sequence is a list of transactions, each of which is a set of literals, called items. Typi-cally, a transaction-time is associated with each transaction. A sequential pattern also con-sists of a list of sets of items. Sequential pattern mining finds all sequential patterns from a time-based transaction database (Agrawal & Srikant, 1995; Srikant & Agrawal, 1996).

The support of an association rule or sequential pattern indicates how frequently the rule applies to the data. A high level of support corresponds to a strong correlation between the product items. The Apriori algorithm (Agrawal et al, 1993; 1994) is typically used to find association rules by discovering frequent itemsets (sets of items). An itemset is considered to be frequent if its support exceeds a user-specified minimum support. Association rules or sequential patterns that meet a user-specified minimum confidence can be generated from the frequent itemsets.

2.5. Context-awareness

According to the definitions of Schilit and Theimer (1994), context is the location of user, the identities of people and objects that are nearby the user, and the status of devices the user interact with. They considered that context-awareness is adapted to the software

(18)

exe-cution environment involving with relevant context changing. Dey et al. (2001) defined the

context as any information that can characterize the situation of an entity, where the entity

can be a user, place, service, and service relevant objects, etc. The context is categorized into location, identity, activity, and time types. A context-aware system uses context to provide relevant information and/or services to the user, where relevancy depends on the user’s task. Ryan et al. (1997) used “environment” to replace “activity” in the context categorization. They use context types to characterize the situation of a particular entity, and provide the information of who, what, when, and where of a particular entity. This work considers the context as any information that can characterize the status of an entity. An entity may be the staff, location, time, or object considered relevant to the interaction between the staff and problem-solving process, including the staff, resolving service provider, components which support resolving service in a problem-solving environment.

2.6. Rule inference with certainty factor

Shortliffe et al. (1975) has proposed the method of Certainty Factor (CF) value to de-rive the certainty degree during the inference, as defined in Equation 3.

⎪ ⎪ ⎪ ⎩ ⎪⎪ ⎪ ⎨ ⎧ < → − → = → > → − − → = → = →                                otherwise Y S Y X Conf if Y S Y S Y X Conf Y X CF Y S Y X Conf if Y S Y S Y X Conf Y X CF Y X CF , 0 ) ( ) ( , ) ( ) ( )) ( ( ) ( ) ( ) ( , ) ( 1 ) ( )) ( ( ) ( ) ( (3)

where the CF value is the certainty degree from -1 to 1; value “1” denotes complete certainty; value “-1” denotes complete uncertainty. In this work, X denotes the preceding set; Y is the set that we want to infer its certainty degree. CF(X→Y) is the CF value of rule X→Y. S(Y) is the support of Y. Conf(X→Y) is the confidence of rule X→Y. Based on the CF value of items and inference rules, the inference process follows the rules defined in Equation 4.

)) ( ), ( ( ) (Xi Xj MIN CF Xi CF Xj CF ∧ = )) ( ), ( ( ) (Xi Xj MAX CF Xi CF Xj CF ∨ = )) ( , 0 ( ) ( }) { | (B IF A THEN B CF A B MAX CF A CF = → × })) { | ( }), { | ( ( )

(B MAX CF B IF A THEN B CF B IF C THEN B

CF =

(19)

Chapter 3. Knowledge Requirements of Knowledge Support for

Problem-solving

In this chapter, we describe the knowledge requirements of knowledge support for problem-solving, including the concepts of the problem-solving process, the knowledge requirements for problem-solving, and the context-based knowledge requirements for problem-solving. A wafer manufacturing process in a semiconductor foundry is used to illustrate the proposed approach. The process comprises the following steps: crystal growing, wafer cutting, edge rounding, lapping, etching, polishing, cleaning, final inspection, pack-aging and shipping. The wafer cleaning step mainly uses DI (de-ionized; ultra-pure) water to remove debris left over from the mounting wax and/or polishing agent. A stable water sup-ply system to deliver ultra-pure water for wafer cleaning is therefore vital in semiconductor manufacturing.

3.1. The problem-solving process

In business enterprises, especially the manufacturing industry, various problem situa-tions may occur during the production process; for example, poor production performance, system overload, and low machine utilization. A situation denotes an evaluation point to determine the status (i.e., desirable or undesirable) of a production process. A problem may occur if there is a discrepancy between the actual situation and the desired one. For example, when the current production output is below the desired level, the production line may have some problems. Thus, a problem-solving process is often initiated to achieve the desired situation. In the process, workers take several problem-solving steps to determine what ac-tion needs to be taken to resolve the situaac-tion. Such acac-tion involves both human wisdom and enterprise knowledge. Workers may observe a problem situation, collect relevant informa-tion from the enterprise knowledge repository, explore possible causes, and identify opera-tional conditions in order to decide appropriate action. Moreover, a problem-solving process generally consists of levels of progressive sub-problem solving, which form different stages of the process. Such stage-wise problem-solving reduces the complexity of a problem and solves it more effectively. The stages of problem-solving in a production process are usually pre-determined by experienced workers or experts according to the characteristics of the process and their experience in solving previous problems.

(20)

3.2. Knowledge requirements for problem-solving

Situation and action relevant knowledge. In a specific stage of problem-solving, a worker

can access relevant documents associated with the problem situation to find the causes. For example, for the situation “crash of the water supply system”, the diagnostic documents contain information about the temperature, pressure, and electric power, which may provide clues to possible causes. The expert-reports indicate that the temperature and pressure fea-tures could be the key reasons for the system’s failure. The experiment-reports show that high pressure may cause an increase in temperature, which would make the system unstable and result in a crash. The know-how hidden in relevant documents can help workers dis-cover the causes of problem situations. These relevant documents are defined as situation relevant knowledge.

After determining the cause of a problem situation, workers must decide what action to take. They do this by accessing documents related to the cause in order to identify the normal operational-conditions of the production system, and choose an appropriate course of action. Continuing with the example of the water system crash, if the cause is an anomalous tem-perature level, a safe temtem-perature range is required to stabilize the system. The system’s operational manual defines the normal pressure and temperature ranges. For example, when the system’s output pressure is one degree of atmospheric pressure, its temperature range is 30 to 32 °C. In addition, the standard operating procedures specify the system’s tuning rules: the system temperature increases 4 °C per degree of atmospheric pressure. The experi-ment-reports indicate a reasonable temperature range of a stable system, where, for example, 55°C is the upper limit of the range. Such relevant operational know-how is hidden in en-terprise documents that must be discovered to help workers take appropriate action, i.e., tune the output pressure and temperature to keep the system stable. These documents are defined as action relevant knowledge.

Decision-making and dependency knowledge. Knowing what action to take to solve

problem situations is defined as decision-making knowledge, which can be discovered from previous problem-solving logs. Decision-making knowledge is expressed as association rules that represent the association of frequently adopted actions for handling specific situations. These knowledge rules are generated as knowledge support to help workers take appropriate action in handling situations. Moreover, in stage-wise problem-solving, a

(21)

three stages of problem-solving on a production line, namely, engineering improvement, quality improvement, and maintenance management.

Engineering Improvement Stage System Crash Situation Tuning Temperature Action Shutting Down Action

Quality Improvemenmt Stage Maintain Management Stage System Control Situation Keeping Working Action Rebooting Action Unstable Quality Situation Evaluating Action

Fig. 1: A problem-solving process for a production line

In the first stage, tuning the system’s temperature and shutting down the system are two appropriate ways to resolve a system crash. The shutting down action may trigger a system control situation, which requires rebooting action in the maintenance management stage. Moreover, the tuning action may cause the situation of unstable quality in the quality im-provement stage. Such cause-effect relationships (chain reactions) across different stages are called dependency knowledge, which helps workers make appropriate action plans across problem-solving stages. Note that decision-making knowledge represents the in-tra-relationships between the situations and actions within a stage, while dependency knowledge denotes the inter-relationships between the situations and actions across different stages.

3.3. Context-based knowledge requirements for problem-solving

Context-based inference. For a given problem, a situation may occur with various features

according to the context at that time. Because situation features collected by system are usually partial or incomplete, a worker can not easily identify current situation. Accordingly, inferring more situation features according to the context characteristics is important in situation identification. For example, the water supply system in a production line provides pure water for wafer cleaning. When the system gets the situation feature “Produc-tion-quality low”, the causes may be so many that a worker can not easily identify the situation. Situation feature “Parameters of water supply quantity service incorrect” is in-ferred from the situation feature “Production quality low” and context characteristic “Pressure of water unstable”. The context characteristic “Pressure of water unstable” and inferred situation features “Parameters of water supply quantity service incorrect” provide CBR with more clues to identify current situation as “Water supply abnormal issue”.

(22)

Context-based situation profile. For specific situation, Information Retrieval (Automatic

Indexing) techniques are used to extract key terms from situation relevant documents. The extracted key terms form a profile to represent the information needs of workers for handling the situation. Moreover, the profile can be generated according to the context of the situation and is regarded as a context-based situation profile. According to certain context, the key terms recorded in a context-based situation profile are used to locate the relevant documents. The relevant documents are recommended as knowledge support to help workers take ap-propriate action for handling the situation in certain context.

Context-based decision-making and dependency knowledge. Knowing what action to take

according to problem situation features is defined as context-based decision-making knowledge, which can be discovered from the problem-solving logs. The context-based decision-making knowledge patterns indicate the inferred associations of actions and situa-tion features in certain context of the problem-solving process. These context-based knowledge patterns are generated as knowledge support to help workers take appropriate action in handling situations. Moreover, in stage-wise problem-solving, a situation/action may trigger/affect a situation/action in a later stage. Context-based dependency knowledge indicates the inferred relationships between situation/action features in current stage and situations/actions across different stages of the whole problem-solving process context. Context-based dependency knowledge helps workers make appropriate action plans across problem-solving stages.

(23)

Chapter 4. Knowledge Support based on Case-based Reasoning

and Data Mining

In this chapter, we describe the knowledge support based on CBR and data mining techniques, including the proposed system framework, discovery of problem-solving knowledge, knowledge support for problem-solving, and a prototype system implementa-tion.

4.1. Knowledge support framework for problem-solving

The proposed knowledge support framework for problem-solving, shown in Fig. 2, employs mining techniques to discover needed knowledge. The system framework com-prises a problem-solving process, knowledge discovery, and knowledge recommendation modules. Knowledge document recommendation Knowledge support network construction Enterprise Knowledge Bases Problem-solving process execution

Knowledge pattern and situation/action profiles

discovery

Knowledge discovery module

Knowledge recommendation module

Employee

Problem-solving process Identification (Case-Based Reasoning)

Intranet Portal

Data processing for knowledge discovery Historical log of intranet

portal

Problem-solving process module

Fig. 2: Knowledge support framework for problem-solving

The proposed framework records the problem-solving steps, including the situations and actions as well as the corresponding knowledge documents accessed in the historical log. The knowledge discovery module employs mining technology to extract hidden knowledge from the historical problem solving log. The extracted knowledge, including situation/action

(24)

profiles, decision-making, and dependency knowledge, is used to provide knowledge sup-port. The knowledge base comprises historical logs, discovered knowledge patterns, situa-tion/action profiles, and enterprise knowledge documents. This component acts as an in-formation hub to provide knowledge support for problem-solving.

Problem-solving process module. This module gathers production run-time information,

such as problem situations. CBR is used to retrieve similar situation/action cases. This is described in Section 4.2. The system then suggests relevant documents and possible knowledge patterns related to the retrieved similar cases. Workers can then execute a spe-cific problem-solving process and obtain knowledge support from the knowledge recom-mendation module. The problem-solving steps, including the situations, actions, and cor-responding knowledge documents accessed, are recorded in the historical log.

Knowledge discovery module. This module searches the historical log file to discover

situation/action profiles and knowledge patterns. The following gives an overview of the knowledge discovery module. Further details are presented in Section 4.2.

z Discovering situation/action profiles. For specific situations or actions, relevant in-formation (documents) accessed by workers is recorded in the problem-solving log. Historical codified knowledge (textual documents) can also provide valuable knowl-edge for solving the target problem. Information Retrieval (Automatic Indexing) tech-niques are used to extract the key terms of relevant documents for a specific situation or action. The extracted key terms form the situation/action profile, which is used to model the information needs of the workers. The knowledge support system then uses the profile to gather relevant information and help workers solve the target problem. Note that relevant information about a situation/action may vary due to a change of enterprise environment. The situation/action profiles can be used to gather existing and new relevant knowledge documents for a specific situation/action.

z Discovering decision-making and dependency knowledge. We assume that a generic problem-solving process is specified by experts to solve a problem or a set of similar problems encountered on a production line. When the production line encounters a problem, a problem-solving process is initiated. The situations occurred in a problem may vary due to the uncertainty of the constantly changing business environment. Moreover, different workers may take different actions to solve a problem according to

(25)

instances. The problem-solving process consists of several stages. Association rule

mining is used to discover decision-making knowledge patterns (intra-relationships)

hidden in a specific stage. Sequential pattern mining is employed to discover de-pendency knowledge patterns (inter-relationships) between different stages (chain re-action). This work employs the Apriori algorithm to find two kinds of rule patterns: association patterns of decision-making knowledge and sequential patterns of de-pendency knowledge. The discovered rule patterns form the basis of decision-making and dependency knowledge. When a situation or action matches a specific knowledge pattern, the associated situations or actions will be suggested as knowledge support.

Knowledge recommendation module. This module constructs a knowledge support network

based on the discovered knowledge patterns and situation/action profiles. A knowledge support network (KSN) is a conceptual representation of knowledge for a specific prob-lem-solving process. It recommends situation/action relevant documents and deci-sion-making/dependency knowledge as knowledge support. As noted previously, the situa-tion/action profiles are used to gather existing and new relevant knowledge documents for a specific situation/action. The situation relevant documents help determine the cause of a problem, while the action-relevant documents (operating procedures and guidelines) instruct workers how to solve it. The KSN also comprises decision-making and dependency knowledge patterns extracted from the knowledge discovery module, and suggests fre-quently adopted actions for handling the problem situation. Dependency knowledge patterns are suggested to help workers infer possible cause-effect relationships and make appropriate action plans across problem-solving stages. The knowledge patterns and relevant documents provide practical knowledge support to help workers solve problems. Further details are presented in Section 4.3.

4.2. Discovery of problem-solving knowledge

This section describes the procedure of discovering knowledge from historical prob-lem-solving logs, as shown in Fig. 3. To illustrate the proposed approach, we use data from the log file of a semiconductor foundry’s intranet portal, which contains the problem-solving log for handling problems on the production line. The company operates wafer manufac-turing fabs to provide the industry with leading-edge foundry services. The log file records the encountered situation and the action taken at each problem stage. The system also con-tains documents accessed by workers for each situation/action during the problem-solving

(26)

process. The data fields of the log include user data and problem-solving data. User data comprises factory, department, and user-role data. Problem-solving data contains the subject (text description) and attribute values of the situation/action, the stages, and the documents accessed.

Knowledge pattern discovery

Data preprocessing for knowledge discovery

Profile discovery

Decision making knowledge patterns

Knowledge discovery module

Dependency knowledge patterns Association rule mining Sequential pattern mining

Situation profiles Action profiles Key terms extraction

Key concept generation Enterprise knowledge bases

Problem-solving process and stage identification

Situation relevant document collection Problem-solving process

Situation and action identification

Knowledge recommendation module

Data transformation

Action relevant document collection

Historical log

Specific stage Specific process

Fig. 3: The procedures of knowledge discovery process.

4.2.1. Data preprocessing for knowledge discovery

The data preprocessing module performs data cleaning, integration, and transformation for further knowledge discovery. The data cleaning task removes inconsistent data from the historical log. Each textual document is transformed into a term vector, i.e., a feature vector of weighted terms, using the tf-idf approach described in Section 2.3. The term vectors of accessed documents are then used by the profile discovery module to generate situa-tion/action profiles. Furthermore, the data records are preprocessed to determine the prob-lem-solving stages and the subject/attribute values of the situations/actions. The extracted

(27)

lem-solving process, and the term vectors of accessed documents are integrated into the enterprise’s knowledge base.

Problem-solving process and stage identification. The Stage field records the problem

category, problem-solving process, and the stage. For example, “Equipment / Water-supply / Engineering - Improvement” shows that the problem category is “Equipment”; the problem solving process is “Water supply”; and the stage is “Engineering Improvement”. The stage field is extracted from the data record to identify the problem-solving process and its stages.

4.2.2. Situation/action identification and case-based reasoning

Each situation or action is a case that is characterized by a text description and a set of attribute values. The attribute values provide additional features, such as the symptoms of a situation or the standard operating procedures of an action to identify the situation/action case. Both the text description and attribute values contribute to similarity matching and situation/action identification. For historical problem-solving instances, similar situa-tion/action cases are transformed into the same situasitua-tion/action identifier to facilitate the mining of decision-making and dependency knowledge patterns. Moreover, for the target situation/action, namely, the case workers are currently handling, the system identifies an existing case identifier or retrieves similar cases based on CBR. In the following, we de-scribe the steps taken to transform existing cases and how to compute the similarity meas-ures for case-based reasoning.

Extraction of identifying term vectors. The data stored in the Subject field of an existing

case is a text description of the situation/action. For example, Subject: “FAB8D Cu-BSC DI Water flow capacity insufficient issue” is the description of the situation - insufficient water flow capacity. The terms extracted from the subject field are used to identify the situa-tion/action. Note that the terms are extracted using term transformation steps, including case folding, stemming, and stop word removal. We simply extract the terms without considering the term frequency, since the subject field generally contains a short text description. The extracted terms form identifying terms to identify a situation/action case. Moreover, the user needs to provide a text description for the target case, namely, the situation or action which he/she is handling. Similarly, the identifying terms of the target case are extracted from the text description using the term transformation steps. Let Tj be the set of identifying terms extracted from the subject field of a situation/action case Cj. An identifying term vector Crj is created to represent Cj. The weight of a term ti inCrj is defined by Equation 5.

(28)

⎩ ⎨ ⎧ ∈ = otherwise T t if C t w i j i j 0 1 ) , ( . (5)

Equation. 6 defines the similarity value simT(Ck, Cj) of two situation/action cases Ck and Cj based on their text descriptions. The similarity value is derived by computing the cosine

value of the identifying term vectors of Ck and Cj.

j k j k j k j k T C C C C C C C C sim r r r r r r • = =cosine( , ) ) , ( (6)

Similarity value by attribute. An attribute value may be nominal, binary, or numeric. For

numeric attributes, a data discretization process is conducted to transform their values into value ranges or user-defined concept terms (such as low, middle or high). Equation 7 defines the similarity value simA(Ck (attrbx); Cj (attrbx)) of two situation/action cases Ck and Cj, derived according to their values of attribute x; value(Ck (attrbx)) denotes the transformed value of attribute x of Ck, which is calculated by the discretization process.

⎩ ⎨ ⎧ = otherwise 0 ) ) ( value( equals )) ( value( if 1 )) ( ), ( ( k x j x k x j x

A C attrb C attb C attrb C attb

sim (7)

Similarity function for case-based reasoning. Equation 8 defines the similarity function

used to compute the similarity measure between two cases Ck and Cj. The similarity function is modified from Guardati (1998) by considering the cosine measure and attribute discreti-zation.

= + = m x k x j x A x j k T T j

k C w sim C C w sim C attrb C attb

C similarity 1 )) ( ), ( ( ) , ( ) , ( , (8) where simT(Ck, Cj) is the similarity value derived from the identifying term vectors of Ck and

Cj; simA(Ck (attrbx); Cj (attrbx)) is the similarity value obtained from the values of attribute x;

wT is the weight factor for the text description, and wx is the weight given to attribute x. Note

that the summation of wT and all wx is equal to 1.

Transforming existing cases. Similar cases are transformed into the same situation/action

(29)

identify cases with high similarity measures (i.e., similarity(Ck, Cj) > θ). Cases with the same

or high similarity measures are transformed into the same situation/action identifier. The transformation procedure is conducted in an incremental and greedy manner. Assume that r situation identifiers have been created. For each Si of r situation identifiers, one or more

situation cases have been transformed into Si. Ck is the situation case that needs to be

transformed into a situation identifier. Let minsim(Ck, Si) be the minimum similarity(Ck, Cj)

over all Cj that is transformed into Si. The procedure finds a situation identifier Sf such that

minsim(Ck , Sf) is the maximum of minsim(Ck, Si) over all Si (for i = 1 to r). For a situation

case Ck, Ck is transformed into Sf, if minsim(Ck, Sf) is greater than θ; otherwise, Ck is

trans-formed into a new situation identifier. The transformation procedure for action cases is conducted in a similar way. Table 1 lists the situations and actions in each stage of the water supply problem-solving process.

Table 1: Situations/actions in the water supply problem-solving process

Water supply problem-solving process

Situations Actions

[S1] Flow Capacity Abnormal Issue (Subject: Insufficient/Unstable/Overflow) [S2] Supply Quantity Abnormal Issue (Subject: Insufficient/Unstable/Overflow) [S3] Power Supply Abnormal Issue (Subject: Insufficient /Unstable/Excess) [S4] Water Pressure Abnormal Issue (Subject: Insufficient/Unstable/Excess) [S5] Cleaning Quality Abnormal Issue (Subject: Low/Unstable)

[S6] Pipe Abnormal Issue (Subject: Broken/Clogged)

[S7] Controller Temperature Abnormal Issue (Subject: Excess/Unstable)

[A1] Testing based on SOPs [A2] Consult expert information [A3] Modify the configuration [A4] Recycle the material [A5] Monitor the output [A6] Discuss with workers [A7] Report the outcome

Case-based reasoning for a target case. A target case is a situation or action that a worker is

currently handling. After entering a target case Ck of a situation/action, the system identifies

an existing case identifier of Ck or retrieves similar situation/action cases if Ck is a new case.

The similarity measures between the target case and previous cases are computed using Equation 6. The identification procedure is similar to the transformation procedure. Assume there are r situation identifiers. Let minsim(Ck, Si) be the minimum similarity(Ck, Cj) over all

Cj transformed into Si. The procedure finds a situation identifier Sf such that minsim(Ck , Sf)

is the maximum of minsim(Ck, Si) over all Si (for i = 1 to r). An existing situation identifier Sf

is identified if minsim(Ck , Sf) is greater than θ; otherwise, the situation is a new case and the

system assigns a new identifier to it. The case and its identifier are then stored in the knowledge base, and CBR is initiated to retrieve similar cases based on their similarity measures and to suggest possible knowledge related to the similar cases.

(30)

r

4.2.3. Discovery of situation/action profiles

The log records the set of documents accessed for handling a situation/action. For example, Doc_ID: “AF0001C0F25” is a planning report that describes how to deal with the abnormal water quality in the DI water system. DI (de-ionized) water is ultra-pure water used for wafer cleaning in semiconductor manufacturing. The term vectors of the documents are derived using Equation 1, i.e., the tf-idf approach described in Section 2.3.

A situation/action profile is also represented as a term vector (a feature vector of weighted terms), which is derived by analyzing the set of documents accessed for handling the situation/action case. Each document dj is pre-processed and represented as a term vector

. Let D

j

d S denote the set of documents accessed to handle the situation/action CS. A centroid

approach is used to derive the profiling term vectorPrSof CS by averaging the term vectors of

documents in DS. Equation 9 defines the weight of a term ki inPS

r .

=

D d j i S S i

w

k

d

D

C

k

w

(

,

)

1

(

,

)

S j . (9)

Retrieval of situation/action relevant documents. The system recommends/retrieves

rele-vant knowledge documents to help workers solve problems based on the situation/action profiles. The key contents of a codified knowledge document are represented as a term vector. The situation/action profile of a case CS is expressed as a profiling term vectorPrS.

The cosine measure of term vectors, described in Section 2.3, is used to derive the similarity measure. Let be the term vector of document ddrj j. The cosine measure of PrSand dj

r , co-sine(PrS ,drj), is the similarity measure between the situation/action and document dj.

Documents with the top-N similarity measures are selected as relevant documents.

4.2.4. Discovery of knowledge patterns

Generic problem-solving process. Recall that a generic problem-solving process is specified

by experts to solve a problem. The specification includes the stages and their execution order. This work focuses on the execution of a sequence of stages. For example, the generic water supply problem-solving process is “Normal Management Stage (NM Stage) → Engineering

(31)

Improvement Stage (EI Stage) → Exception Management Stage (EM Stage) → Quality Improvement Stage (QI Stage) → Maintenance Management Stage (MM Stage). For any given problem, the situations may vary; thus the follow-up actions may also vary.

Discovery of decision-making knowledge patterns. Association rule mining is used to

discover decision-making knowledge hidden in each problem solving stage. In this paper, we adopt the Apriori algorithm to find the frequent association patterns of decision-making knowledge, namely situation → action. The criteria of minimum support and confidence are used to filter out non-frequent patterns. The discovered rule patterns form the basis of de-cision-making knowledge. When a situation matches a specific knowledge pattern, the as-sociated action will be suggested as knowledge support. For example, the discovered deci-sion-making knowledge patterns in the quality improvement stage (QI) are:

z QI_S1 → QI_A3

If the situation “Water Flow Capacity Abnormal Issue (Insufficient)” occurs, take the “Modify the configuration” action.

z QI_S6 → QI_A5

If the situation “Pipe Abnormal Issue (Clogged)” occurs, take the “Monitor the output” action.

Discovery of dependency knowledge patterns. Sequential pattern mining is adopted to

discover the dependency knowledge patterns (inter-relationships) hidden between stages. We use a modified Apriori algorithm for sequential pattern mining to discover the frequency of similar situations and actions across different stages. The criterion of minimum support is used to filter out non-frequent (chain reaction) relationships. When a situation or action matches a specific knowledge pattern, the chain of situations or actions is suggested as knowledge support. Some examples of dependency knowledge patterns are:

z EI_S1 → QI_S4

If the situation “Flow Capacity Abnormal Issue (Overflow)” occurs in the engi-neering improvement stage, then it is likely that the situation “Water Pressure Abnormal Issue (Excess)” will occur in the quality improvement stage.

(32)

z EI_A4 → EM_S1

If the “Recycle the material” action is taken in the engineering improvement stage, then the situation “Flow Capacity Abnormal Issue (Unstable)” is likely to occur in the exception management stage.

z NM_A3 → QI_A5

If the “Modify the configuration” action is taken in the normal management stage, then the “Monitor the output” action is likely to be taken in the quality improve-ment stage.

z EI_S2 → EM_A7 → QI_S1

The situation “Supply Quantity Abnormal Issue (Insufficient)” in the engineering improvement stage frequently triggers the “Report the outcome” action in the exception management stage; and then triggers the “Flow Capacity Abnormal Issue (Insufficient)” situation in the quality improvement stage.

The dependency knowledge patterns denote the chain reaction across different stages. This helps workers plan appropriate actions for different problem-solving stages. The deci-sion-making and dependency knowledge patterns are integrated into the knowledge support network.

4.3. Knowledge support for problem-solving

This section describes the construction of the knowledge support network, which pro-vides knowledge recommendations for problem-solving. The procedure is showed in Fig. 4.

4.3.1. Knowledge support network

A knowledge support network (KSN) is constructed from the output of the knowledge discovery module. The KSN comprises the specification of the generic problem-solving process, decision-making and dependency knowledge patterns, situation/action profiles, and relevant documents.

Specification of a generic problem-solving process. The specification includes the problem

(33)

Knowledge Discovery Module

Knowledge Support Network

Knowledge Recommendation Module

Knowledge Recommendation Specify/Select the subject of situation/action

Identify the situation/action

Match the decision-making and dependency knowledge patterns Knowledge Support Network Construction

Specific problem-solving

process template Knowledge patterns

Situation/Action profiles Relevant enterprise document set Situation Description Relevant Documents

Specific Problem-Solving Process Stage 1 Action Description Situation Description Stage 2 Action Description Situation Description Stage N Action Description Situation Description Stage N-1 Action Description ... ... Situation Profiles Relevant Documents Situation Profiles Relevant Documents Situation Profiles Relevant Documents Situation Profiles Relevant Documents Action Profiles Relevant Documents Action Profiles Relevant Documents Action Profiles Relevant Documents Action Profiles S S S A A A S S S A A A A A A S S S A A A S S S

Recommend the relevant knowledge documents

Fig. 4: The procedures of knowledge recommendation

Decision-making knowledge patterns. Decision-making knowledge patterns indicate the

frequent association of situations and actions in the problem-solving process. For each stage, a decision-making knowledge pattern situation → action indicates that the action frequently adopted to solve the encountered problem situation. The KSN provides frequently adopted actions for handling a specific situation based on the decision-making knowledge patterns. Fig. 5 shows the discovered decision-making knowledge patterns in the KSN of the water supply problem-solving process.

(34)

W ater supply problem -solving process

N M Stage E I Stage E M Stage

Situation

Q I Stage

Situation

M M Stage

A ction Situation A ction Situation A ction Situation A ction A ction

N M _S7 N M _S6 N M _A 5 N M _A 7 N M _A 3 EI_A 1 E I_A 2 E I_A 4 E I_A 5 E I_S2 E I_S1 EI_S4 EM _S1 EM _S2 EM _S3 EM _S4 EM _A 1 EM _A 7 EM _A 3 Q I_S1 Q I_S6 Q I_A 3 Q I_A 5 M M _A 6 M M _S1 Q I_S4 M M _A 5

Fig. 5: Decision-making knowledge patterns in a knowledge support network

Dependency knowledge patterns. For a specific problem-solving process, the dependency

knowledge patterns express the relationships between situations and actions across different stages. For example, the dependency knowledge pattern “EM_S3 → MM_A5” implies that if a “Power Supply Abnormal Issue (Unstable)” situation occurs in the exception manage-ment stage, then the “Monitor the output” action is frequently taken in the maintenance management stage. A dependency knowledge pattern “EI_S4 → QI_A5→ MM_A6” plies that a “Water Pressure Abnormal Issue (Unstable)” situation in the engineering im-provement stage will trigger a “Monitor the outcome” action in the quality imim-provement stage; and then trigger a “Discuss with the worker” action in the maintenance management stage. Based on the dependency knowledge patterns, the KSN provides triggering situations or actions across different stages, which help workers predict possible situations in later stages, and plan appropriate actions. Fig. 6 shows the dependency knowledge patterns in a KSN.

(35)

Situation/action profiles and relevant documents. The situation/action profiles are

gener-ated from the accessed documents, as described in section 4.2. For example, in the situation of abnormal water quantity, the accessed documents include: “DI analytical machine water quantity recording” and “DI GCHC machine water quantity recycling.” The situation profile is generated from the accessed documents. Once a worker encounters a problem situation or decides to take a particular action, the KSN provides relevant documents as knowledge support based on the situation/action profiles. Fig. 7 illustrates a situation profile and the relevant documents for the water supply problem-solving process. Based on the situa-tion/action profiles, the knowledge support network gathers previous and new relevant documents, such as “DI analytical machine water quantity recording” and “DI GCHC ma-chine water quantity recycling” and new documents “8D DI system waste water quantity estimation” and “8D UF Flush water quantity recycling”.

4.3.2. Knowledge recommendation

The problem-solving process module employs CBR to identify the current situation or retrieve similar situation-cases according to the similarity measures. The knowledge rec-ommendation module then suggests relevant documents according to the situation profile of the current situation or similar cases, as shown in Fig. 7. The system also recommends relevant action documents (e.g., operating procedures and guidelines) according to the ac-tion profile. Note that the top-N relevant documents are recommended according to the co-sine measure of the term vectors of the documents and the situation/action profiles, as de-scribed in Section 4.2.

Moreover, the system suggests possible actions for handling the current situation ac-cording to the decision-making knowledge patterns. Note that the actions in the deci-sion-making patterns (i.e., situation => action) whose left-hand side match the current situation are suggested and ranked according to the confidence values of the rules. De-pendency knowledge patterns are also suggested to help workers predict a possible chain reaction across different stages and develop appropriate action plans.

(36)

Fig. 7: Situation profile and relevant documents

4.4. System implementation

We developed a prototype system to demonstrate the effectiveness of the proposed knowledge support system for problem-solving. The implementation is conducted using several software tools, including the Java(TM) 2 Platform Standard Edition Runtime En-vironment Version 5.0, Java Server Page, and Macromedia Dreamweaver MX. A web and application server is setup on Apache Tomcat 5.5.7, and Microsoft SQL Server 2000 is used as the database system for storing data related to the problem-solving process and codified knowledge documents. The data mining tool Weka 3.4 is used to discover knowledge pat-terns in the historical problem-solving log.

The generic problem-solving process, situation/action profiles, decision-making and dependency knowledge patterns form the knowledge support network. The network pro-vides relevant knowledge documents, and suggests decision-making and dependency knowledge patterns. The problem-solving knowledge support system is integrated with the knowledge support network to provide more effective knowledge support for browsing problem-solving knowledge patterns. The interface of the problem-solving knowledge support system includes the system frames for user login, search engine, and user-guide. A worker Annie logs into the system and gets a problem list. Once she selects a generic problem-solving process to browse, the problem (e.g., water supply problem) can be browsed further in the system platform, as shown in Fig. 8.

數據

Table 1: Situations/actions in the water supply problem-solving process..........................
Fig. 1: A problem-solving process for a production line
Fig. 2: Knowledge support framework for problem-solving
Fig. 3: The procedures of knowledge discovery process.  4.2.1. Data preprocessing for knowledge discovery
+7

參考文獻

相關文件

&#34;Extensions to the k-Means Algorithm for Clustering Large Data Sets with Categorical Values,&#34; Data Mining and Knowledge Discovery, Vol. “Density-Based Clustering in

This kind of algorithm has also been a powerful tool for solving many other optimization problems, including symmetric cone complementarity problems [15, 16, 20–22], symmetric

In order to solve the problems mentioned above, the following chapters intend to make a study of the structure and system of The Significance of Kuangyin Sūtra, then to have

In addition to examining the influence that the teachings of Zen had on Shi Tao’s art and theoretical system, this paper proposes further studies on Shi Tao’s interpretation on

4. To apply the basic principles and techniques in preparing personal budget, and 5. To develop a proper attitude towards personal finance.. Resources for the TEKLA curriculum

Referring to the Student Worksheet (p.2) of Activity 2, ask students to identify which situations incur pure risk and speculative risk respectively... Invite students to present

Since the FP-tree reduces the number of database scans and uses less memory to represent the necessary information, many frequent pattern mining algorithms are based on its

In this chapter, we have presented two task rescheduling techniques, which are based on QoS guided Min-Min algorithm, aim to reduce the makespan of grid applications in batch