1. 10.1 c h a p t e r 10 MANAGING KNOWLEDGE: KNOWLEDGE WORK AND ARTIFICIAL INTELLIGENCE

2. 10.2 LEARNING OBJECTIVES EXPLAIN ORGANIZATIONAL KNOWLEDGE MANAGEMENTEXPLAIN ORGANIZATIONAL KNOWLEDGE MANAGEMENT DESCRIBE USEFUL APPLICATIONS FOR DISTRIBUTING, CREATING, SHARING KNOWLEDGEDESCRIBE USEFUL APPLICATIONS FOR DISTRIBUTING, CREATING, SHARING KNOWLEDGE EVALUATE ROLE OF ARTIFICIAL INTELLIGENCE IN KNOWLEDGE MANAGEMENTEVALUATE ROLE OF ARTIFICIAL INTELLIGENCE IN KNOWLEDGE MANAGEMENT

3. 10.3 LEARNING OBJECTIVES DEMONSTRATE HOW ORGANIZATIONS USE EXPERT SYSTEMS, CASE-BASED REASONING TO CAPTURE KNOWLEDGEDEMONSTRATE HOW ORGANIZATIONS USE EXPERT SYSTEMS, CASE-BASED REASONING TO CAPTURE KNOWLEDGE

4. 10.4 QUAKER CHEMICAL Quaker Chemical, based, Pennsylvania, is a worldwide producer of custom-formulated chemical specialty products and fluid management services.Quaker Chemical, based, Pennsylvania, is a worldwide producer of custom-formulated chemical specialty products and fluid management services. Clients:Clients: –Makers of automotive –Aerospace products – Firms in the environment –Can industries –Pulp industries and paper industries.

5. 10.5 QUAKER CHEMICAL Recently underwent a global reorganization so that its sales staff now sells by business line rather than by geographic region.

6. 10.6 QUAKER CHEMICAL Quaker needed a system that could rise above regional barriers and help its sales force:Quaker needed a system that could rise above regional barriers and help its sales force: – Track, manage, and collaborate on client services and accounts

7. 10.7 QUAKER CHEMICAL Developing new chemical formulas for automakersDeveloping new chemical formulas for automakers Determining why sales at certain North American mills are down.Determining why sales at certain North American mills are down. These problems could be more easily solved if the company could gather information stored on workers' computers and in their heads and make it more widely available. These problems could be more easily solved if the company could gather information stored on workers' computers and in their heads and make it more widely available.

8. 10.8 QUAKER CHEMICAL Quaker doesn't have off-the-shelf products.Quaker doesn't have off-the-shelf products. Quaker formulates unique products for every customer.Quaker formulates unique products for every customer. capturing what Quaker's laboratory workers know about Quaker formulas is especially critical.

9. 10.9 QUAKER CHEMICAL In mid-2000, Quaker launched a KNOWLEDGE MANAGEMENT SYSTEM called Quaker Business Intelligence, or QBI.In mid-2000, Quaker launched a KNOWLEDGE MANAGEMENT SYSTEM called Quaker Business Intelligence, or QBI. The system is a global intranet with collaborative software from Intraspect Software of Brisbane, California, working in conjunction with Quaker's Novell GroupWise collaboration systemThe system is a global intranet with collaborative software from Intraspect Software of Brisbane, California, working in conjunction with Quaker's Novell GroupWise collaboration system

10. 10.10 QUAKER CHEMICAL Employees can DROP word processing documents, e-mail, Web pages, presentations, and spreadsheets in CENTRAL FILES into the system.Employees can DROP word processing documents, e-mail, Web pages, presentations, and spreadsheets in CENTRAL FILES into the system. They can SUBSCRIBE to certain folders relating to their jobs in the same way they would subscribe to an e-mail mailing list.They can SUBSCRIBE to certain folders relating to their jobs in the same way they would subscribe to an e-mail mailing list. The system ALERTS them automatically when changes are made to the individual files RELATING TO THEIR JOB tasks.The system ALERTS them automatically when changes are made to the individual files RELATING TO THEIR JOB tasks.

11. 10.11 QUAKER CHEMICAL When an employee LOGS on to his or her computer, the Intraspect software POPS UP with the look and feel of a simple Web page. Employees can CONTROL what is added to their "HOME PAGE" or to the files belonging to their group and they can customize their home page to fit their job needs.When an employee LOGS on to his or her computer, the Intraspect software POPS UP with the look and feel of a simple Web page. Employees can CONTROL what is added to their "HOME PAGE" or to the files belonging to their group and they can customize their home page to fit their job needs. The Intraspect system can be INTEGRATED with a data warehouse that Quaker uses to track its FINANCIAL INFORMATION.The Intraspect system can be INTEGRATED with a data warehouse that Quaker uses to track its FINANCIAL INFORMATION.

12. 10.12 QUAKER CHEMICAL About 800 of Quaker's 1,100 WORKERS use the Intraspect system.About 800 of Quaker's 1,100 WORKERS use the Intraspect system. According to Thomas Baker, Quaker's manager of business intelligence development, the system stores nearly 27,000 documents in over 7,000 folders.According to Thomas Baker, Quaker's manager of business intelligence development, the system stores nearly 27,000 documents in over 7,000 folders.

13. 10.13 QUAKER CHEMICAL PREVIOUSLY, if a manager had trouble finding out why a customer experienced SURFACE IMPERFECTIONS while rolling steel in a specific type of mill, he or she would E-MAIL OTHER QUAKER COLLEAGUES.PREVIOUSLY, if a manager had trouble finding out why a customer experienced SURFACE IMPERFECTIONS while rolling steel in a specific type of mill, he or she would E-MAIL OTHER QUAKER COLLEAGUES. NOW that manager can INITIATE A DISCUSSION within QBI, sending an alert to all managers in the company's steel division.NOW that manager can INITIATE A DISCUSSION within QBI, sending an alert to all managers in the company's steel division. The system archives the message threads from that discussion and these threads can be ACCESSED LATER by workers trying to answer a SIMILAR QUESTION.The system archives the message threads from that discussion and these threads can be ACCESSED LATER by workers trying to answer a SIMILAR QUESTION.

14. 10.14 QUAKER CHEMICAL Many of these benefits are DIFFICULT TO QUANTIFY.Many of these benefits are DIFFICULT TO QUANTIFY. However, Quaker claims that QBI has already saved the company FOUR MONTHS of labor worth $300,000 by enabling THREE DIFFERENT LAB SITES to access formulas stored in the system rather than DUPLICATE the research.However, Quaker claims that QBI has already saved the company FOUR MONTHS of labor worth $300,000 by enabling THREE DIFFERENT LAB SITES to access formulas stored in the system rather than DUPLICATE the research.

15. 10.15 MANAGEMENT CHALLENGES Designing knowledge systems that GENUINELY ENHANCE organizational performance.Designing knowledge systems that GENUINELY ENHANCE organizational performance. Information systems that truly enhance the productivity of knowledge workers may be difficult to build because the manner in which information technology can enhance higher-level tasks, such as those performed by MANAGERS and professionals, is NOT ALWAYS CLEARLY UNDERSTOOD.Information systems that truly enhance the productivity of knowledge workers may be difficult to build because the manner in which information technology can enhance higher-level tasks, such as those performed by MANAGERS and professionals, is NOT ALWAYS CLEARLY UNDERSTOOD.

16. 10.16 MANAGEMENT CHALLENGES Some aspects of organizational knowledge cannot be CAPTURED EASILY OR CODIFIED.Some aspects of organizational knowledge cannot be CAPTURED EASILY OR CODIFIED. The information that organizations finally manage to capture may become OUTDATED as environments change.The information that organizations finally manage to capture may become OUTDATED as environments change. It is VERY DIFFICULT to INTEGRATE knowledge management programs with business strategy.It is VERY DIFFICULT to INTEGRATE knowledge management programs with business strategy.

17. 10.17 MANAGEMENT CHALLENGES –IDENTIFYING and implementing APPROPRIATE organizational applications for ARTIFICIAL INTELLIGENCE. –Only CERTAIN KINDS of information problems are APPROPRIATE for artificial intelligence (AI) applications. –Many AI applications IMPROVE performance through TRIAL AND ERROR and may not be reliable enough for MISSION-CRITICAL PROBLEMS.

18. 10.18 MANAGEMENT CHALLENGES –EXPERT SYSTEMS are EXPENSIVE and TIME-CONSUMING to maintain because their rules must be reprogrammed every time there is a change in the organizational environment. –MANY THOUSANDS of businesses have undertaken experimental projects in expert systems, but ONLY A SMALL percentage have created expert systems that ACTUALLY CAN BE USED ON A PRODUCTION BASIS.

19. 10.19

20. 10.20 KNOWLEDGE MANAGEMENT IN THE ORGANIZATION As knowledge becomes a CENTRAL PRODUCTIVE and STRATEGIC ASSET. ORGANIZATIONAL SUCCESS increasingly DEPENDS on the firm's ability to:As knowledge becomes a CENTRAL PRODUCTIVE and STRATEGIC ASSET. ORGANIZATIONAL SUCCESS increasingly DEPENDS on the firm's ability to: –PRODUCE KNOWLEDGE –GATHER KNOWLEDGE –STORE KNOWLEDGE –DISSEMINATE KNOWLEDGE. –With KNOWLEDGE, firms become more EFFICIENT and EFFECTIVE in their use of scarce resources.

21. 10.21 55 percent of the U.S. labor force consists of knowledge and information workers.55 percent of the U.S. labor force consists of knowledge and information workers. Producing unique products or services or producing them at a lower cost than competitors is based on SUPERIOR KNOWLEDGE of the production process and superior design.Producing unique products or services or producing them at a lower cost than competitors is based on SUPERIOR KNOWLEDGE of the production process and superior design. Knowing how to do things effectively and efficiently in ways that other organizations cannot duplicate is a PRIMARY SOURCE OF VALUE and a factor in production that cannot be purchased in external markets.Knowing how to do things effectively and efficiently in ways that other organizations cannot duplicate is a PRIMARY SOURCE OF VALUE and a factor in production that cannot be purchased in external markets. KNOWLEDGE MANAGEMENT IN THE ORGANIZATION

22. 10.22 Some management theorists believe that these KNOWLEDGE ASSETS are as important as the PHYSICAL and FINANCIAL assets.Some management theorists believe that these KNOWLEDGE ASSETS are as important as the PHYSICAL and FINANCIAL assets. KNOWLEDGE MANAGEMENT IN THE ORGANIZATION

23. 10.23 KNOWLEDGE MANAGEMENT refers to the SET OF PROCESSES developed in an organization to CREATE, GATHER, STORE, TRANSFER, and APPLY knowledge.KNOWLEDGE MANAGEMENT refers to the SET OF PROCESSES developed in an organization to CREATE, GATHER, STORE, TRANSFER, and APPLY knowledge. INFORMATION TECHNOLOGY plays an important role in knowledge management by SUPPORTING THESE BUSINESS PROCESSES for creating, identifying, and leveraging knowledge throughout the organization.INFORMATION TECHNOLOGY plays an important role in knowledge management by SUPPORTING THESE BUSINESS PROCESSES for creating, identifying, and leveraging knowledge throughout the organization. KNOWLEDGE MANAGEMENT IN THE ORGANIZATION

24. 10.24 Organizations create and gather knowledge through a variety of ORGANIZATIONAL LEARNING mechanisms.Organizations create and gather knowledge through a variety of ORGANIZATIONAL LEARNING mechanisms. Through:Through: TRIAL and ERRORTRIAL and ERROR CAREFUL MEASUREMENT of planned activities.CAREFUL MEASUREMENT of planned activities. FEEDBACK from customers and the environment in general.FEEDBACK from customers and the environment in general. Organizations create new SOPs and business processes that reflect their experience KNOWLEDGE MANAGEMENT IN THE ORGANIZATION

25. 10.25 CHIEF KNOWLEDGE OFFICER (CKO) DIGITAL FIRM: The chief knowledge officer is a SENIOR EXECUTIVE who is responsible for the FIRM'S KNOWLEDGE MANAGEMENT PROGRAM.The chief knowledge officer is a SENIOR EXECUTIVE who is responsible for the FIRM'S KNOWLEDGE MANAGEMENT PROGRAM.

26. 10.26 This knowledge base may include:This knowledge base may include: 1.Structured, INTERNAL KNOWLEDGE (explicit knowledge), such as product manuals or research reports 2.EXTERNAL KNOWLEDGE of competitors, products, and markets, including competitive intelligence 3.INFORMAL, internal knowledge, often called TACIT KNOWLEDGE, which resides in the minds of individual employees but has not been documented in structured form Systems and Infrastructure for Knowledge Management

27. 10.27 Companies can use information systems to codify their BEST PRACTICES and make knowledge of these practices widely available to employees BEST PRACTICES Successful solutions or problem-solving methods developed by specific organization or industry Systems and Infrastructure for Knowledge Management

28. 10.28 The knowledge can be PRESERVED as ORGANIZATIONAL MEMORY to TRAIN FUTURE employees or to help them with decision makingThe knowledge can be PRESERVED as ORGANIZATIONAL MEMORY to TRAIN FUTURE employees or to help them with decision making ORGANIZATIONAL MEMORY Stored learning from organization’s historyStored learning from organization’s history Used for decision making and other purposesUsed for decision making and other purposes Systems and Infrastructure for Knowledge Management

29. 10.29 KNOWLEDGE MANAGEMENT IN THE ORGANIZATION KNOWLEDGE MANAGEMENT: Office Automation Systems (OAS) Knowledge Work Systems (KWS) Group Collaboration Systems (GCS) Artificial Intelligence Applications (AI) *

30. 10.30

31. 10.31 Information Work is work that consists primarily of CREATING OR PROCESSING INFORMATION.Information Work is work that consists primarily of CREATING OR PROCESSING INFORMATION. It is carried out by Information workers who usually are divided into two subcategories: It is carried out by Information workers who usually are divided into two subcategories: 1.DATA WORKERS, who primarily process and disseminate information 2.KNOWLEDGE WORKERS, who primarily create knowledge and information. INFORMATION AND KNOWLEDGE WORK SYSTEMS

32. 10.32 Most data work and a great deal of knowledge work takes place in offices, using office systems including most of the work done by managersMost data work and a great deal of knowledge work takes place in offices, using office systems including most of the work done by managers Distributing Knowledge: Office and Document Management Systems

33. 10.33 Office systems Manage and coordinate work of data and knowledge workersManage and coordinate work of data and knowledge workers Connect work of local information workers with all levels and functions of organizationConnect work of local information workers with all levels and functions of organization Connect organization to external worldConnect organization to external world Example: Word processing, voice mail, and imagingExample: Word processing, voice mail, and imaging Distributing Knowledge: Office and Document Management Systems

34. 10.34 INFORMATION AND KNOWLEDGE WORK SYSTEMS The Three Major Roles of Offices Figure 10-2

35. 10.35 OFFICE WORKERS Office workers span a very broad range: professionals, managers, sales, and clerical workers working alone or in groups. Their major activities include the following:

36. 10.36 OFFICE AUTOMATION SYSTEMS MANAGING DOCUMENTS: CREATIONCREATION STORAGESTORAGE RETRIEVALRETRIEVAL DISSEMINATIONDISSEMINATION TECHNOLOGY: Word processing, desktop publishing, document imaging, Web publishing, work flow managersTECHNOLOGY: Word processing, desktop publishing, document imaging, Web publishing, work flow managers*

37. 10.37 OFFICE AUTOMATION SYSTEMS SCHEDULING: FOR INDIVIDUALS & GROUPS: ELECTRONIC CALENDARSELECTRONIC CALENDARS GROUPWAREGROUPWARE INTRANETSINTRANETS

38. 10.38 OFFICE AUTOMATION SYSTEMS COMMUNICATING: INITIATING, RECEIVING, MANAGING: VOICEVOICE DIGITALDIGITAL DOCUMENTSDOCUMENTS TECHNOLOGY: E-mail, voice mail, digital answering systems, GroupWare, intranetsTECHNOLOGY: E-mail, voice mail, digital answering systems, GroupWare, intranets*

39. 10.39 OFFICE AUTOMATION SYSTEMS MANAGING DATA: EMPLOYEES, CUSTOMERS, VENDORS: 1.DESKTOP DATABASES 2.SPREADSHEETS 3.USER-FRIENDLY INTERFACES TO MAINFRAME DATABASES

40. 10.40 DOCUMENT IMAGING SYSTEMS: Systems convert documents, images into digital form (e.g.: optical character recognition; microfiche)DOCUMENT IMAGING SYSTEMS: Systems convert documents, images into digital form (e.g.: optical character recognition; microfiche) JUKEBOX: Storage & retrieving device for CD-ROMs & other optical disksJUKEBOX: Storage & retrieving device for CD-ROMs & other optical disks INDEX SERVER: Imaging system to store / retrieve documentINDEX SERVER: Imaging system to store / retrieve document* OFFICE AUTOMATION SYSTEMS : OFFICE AUTOMATION SYSTEMS Document Management :

41. 10.41 INFORMATION AND KNOWLEDGE WORK SYSTEMS Components of an Imaging System Figure 10-3

42. 10.42 INFORMATION AND KNOWLEDGE WORK SYSTEMS Web Publishing and Document Management Figure 10-4

43. 10.43 CREATE KNOWLEDGE KNOWLEDGE WORK SYSTEMS: INFORMATION SYSTEMS THAT AID KNOWLEDGE WORKERS TO CREATE, INTEGRATE NEW KNOWLEDGE IN ORGANIZATION INFORMATION SYSTEMS THAT AID KNOWLEDGE WORKERS TO CREATE, INTEGRATE NEW KNOWLEDGE IN ORGANIZATION

44. 10.44 REQUIREMENTS OF KNOWLEDGE WORK SYSTEMS

45. 10.45 CREATE KNOWLEDGE KNOWLEDGE WORKERS: 1.KEEP ORGANIZATION UP-TO-DATE IN KNOWLEDGE: Technology; science; thought; the arts 2.INTERNAL CONSULTANTS IN THEIR AREAS. 3.CHANGE AGENTS: Evaluating; initiating; promoting change projects

46. 10.46 CAD/CAM Computer Aided Design/Computer Aided Manufacturing: Provides precise control over industrial design, manufacturingComputer Aided Design/Computer Aided Manufacturing: Provides precise control over industrial design, manufacturing VIRTUAL REALITY VIRTUAL REALITY Interactive software creates photorealistic simulations of real world objects (Virtual Reality Modeling Language: VRML) Interactive software creates photorealistic simulations of real world objects (Virtual Reality Modeling Language: VRML) CREATE KNOWLEDGE EXAMPLES KNOWLEDGE SYSTEMS

47. 10.47 INVESTMENT WORKSTATIONS investment workstations is High-end PCs used in finance to analyze trading situations, that integrate a wide range of data from both internal and external sources, including contact management data, real-time and historical market data, and research reports.investment workstations is High-end PCs used in finance to analyze trading situations, that integrate a wide range of data from both internal and external sources, including contact management data, real-time and historical market data, and research reports. Previously, financial professionals had to spend considerable time accessing data from separate systems and piecing together the information they needed.Previously, financial professionals had to spend considerable time accessing data from separate systems and piecing together the information they needed. By providing one-stop information faster and with fewer errors. CREATE KNOWLEDGE EXAMPLES KNOWLEDGE SYSTEMS

48. 10.48 SHARE KNOWLEDGE GROUP COLLABORATION SYSTEMS: GROUPWARE: Allows interactive collaboration.GROUPWARE: Allows interactive collaboration. Communication, Collaboration, and coordination.Communication, Collaboration, and coordination. It allows groups to work together on: It allows groups to work together on: Documents Documents Schedule meetings, Schedule meetings, Access shared folders, Access shared folders, Participate in electronic discussionsParticipate in electronic discussions Develop shared databasesDevelop shared databases Send e-mail.Send e-mail.

49. 10.49 SHARE KNOWLEDGE GROUP COLLABORATION SYSTEMS: INTRANETS: Good for relatively stable information in central repository.INTRANETS: Good for relatively stable information in central repository. Ford Motor CompanyFord Motor Company Intranet delivers information about news, people, processes, products, and competition to 95,000 professional employees.Intranet delivers information about news, people, processes, products, and competition to 95,000 professional employees. Employees can access online libraries and a Web Center of Excellence with information on best practices, standards, and recommendations.Employees can access online libraries and a Web Center of Excellence with information on best practices, standards, and recommendations.

50. 10.50 Figure 10-7 ENTERPRISE INFORMATION PORTALS

51. 10.51 TEAMWARE: Consists of intranet-based applications for:TEAMWARE: Consists of intranet-based applications for: Building a work teamBuilding a work team Sharing ideas and documentsSharing ideas and documents BrainstormingBrainstorming Scheduling, tracking the status of tasks and projectsScheduling, tracking the status of tasks and projects Teamware is similar to groupware, although its application development capabilities are not as powerful as those provided by sophisticated groupware products.Teamware is similar to groupware, although its application development capabilities are not as powerful as those provided by sophisticated groupware products. SHARE KNOWLEDGE GROUP COLLABORATION SYSTEMS:

52. 10.52 CAPABILITIES OF GROUPWARE PUBLISHING, REPLICATIONPUBLISHING, REPLICATION DISCUSSION TRACKINGDISCUSSION TRACKING DOCUMENT MANAGEMENTDOCUMENT MANAGEMENT WORK-FLOW MANAGEMENTWORK-FLOW MANAGEMENT PORTABILITYPORTABILITY APPLICATION DEVELOPMENTAPPLICATION DEVELOPMENT

53. 10.53 ARTIFICIAL INTELLIGENCE (AI) SYSTEMS: AI: COMPUTER-BASED SYSTEMS WITH ABILITIES TO LEARN LANGUAGE, ACCOMPLISH TASKS, USE PERCEPTUAL APPARATUS, EMULATE HUMAN EXPERTISE & DECISION MAKING * AI

54. 10.54 AI FAMILY NATURAL LANGUAGE ROBOTICS PERCEPTIVE SYSTEMS EXPERT SYSTEMS INTELLIGENT MACHINES ARTIFICIAL INTELLIGENCE AI

55. 10.55 BUSINESS INTERESTS IN AI 1.PRESERVE EXPERTISE 2.CREATE KNOWLEDGE BASE 3.MECHANISM NOT SUBJECT TO FEELINGS, FATIGUE, WORRY, CRISIS 4.ELIMINATE ROUTINE / UNSATISFYING JOBS 5.ENHANCE KNOWLEDGE BASE (generating solutions to specific problems that are too massive and complex to be analyzed by human beings in a short period of time) *

56. 10.56 Rules in an AI Program ARTIFICIAL INTELLIGENCE Figure 10-9

57. 10.57 EXPERT SYSTEMS EXPERT SYSTEMS KNOWLEDGE - INTENSIVE CAPTURES HUMAN EXPERTISE IN LIMITED DOMAINS OF KNOWLEDGE * AI

58. 10.58 EXPERT SYSTEMS EXPERT SYSTEMS KNOWLEDGE BASE: Model of Human KnowledgeKNOWLEDGE BASE: Model of Human Knowledge RULE - BASED EXPERT SYSTEM : AI system based on IF - THEN statements (Bifurcation); Rule Base: Collection of IF - THEN knowledgeRULE - BASED EXPERT SYSTEM : AI system based on IF - THEN statements (Bifurcation); Rule Base: Collection of IF - THEN knowledge KNOWLEDGE FRAMES: Knowledge organizes in chunks based on shared relationshipsKNOWLEDGE FRAMES: Knowledge organizes in chunks based on shared relationships* AI

59. 10.59 EXPERT SYSTEMS EXPERT SYSTEMS AI SHELL: Programming environment of expert systemAI SHELL: Programming environment of expert system INFERENCE ENGINE: Search through rule baseINFERENCE ENGINE: Search through rule base –FORWARD CHAINING: –FORWARD CHAINING: Uses input; searches rules for answer –BACKWARD CHAINING: –BACKWARD CHAINING: Begins with hypothesis, seeks information until hypothesis accepted or rejected (Should we add this person to the prospect database?) AI

60. 10.60 ARTIFICIAL INTELLIGENCE Figure 10-10

61. 10.61 Building an Expert System The team members must select a problem appropriate for an expert system.The team members must select a problem appropriate for an expert system. The project will balance potential savings from the proposed system against the cost.The project will balance potential savings from the proposed system against the cost. The team members will develop a prototype system to test assumptions about how to encode the knowledge of experts.The team members will develop a prototype system to test assumptions about how to encode the knowledge of experts. Next, they will develop a full-scale system, focusing mainly on the addition of a very large number of rules.Next, they will develop a full-scale system, focusing mainly on the addition of a very large number of rules.

62. 10.62 Building an Expert System The system will be pruned to achieve simplicity and power.The system will be pruned to achieve simplicity and power. The system is tested by a range of experts within the organization against the performance criteria established earlier.The system is tested by a range of experts within the organization against the performance criteria established earlier. Once tested, the system will be integrated into the data flow and work patterns of the organization.Once tested, the system will be integrated into the data flow and work patterns of the organization.

63. 10.63 Knowledge engineer Specialist extract information and expertise from other professionalsSpecialist extract information and expertise from other professionals Translates information into set of rules for an expert systemTranslates information into set of rules for an expert system Building an Expert System

64. 10.64 Countrywide Funding Corp. is a loan-underwriting firm with about 400 underwriters in 150 offices around the country.Countrywide Funding Corp. is a loan-underwriting firm with about 400 underwriters in 150 offices around the country. The company developed a PC-based expert system in 1992 to make preliminary creditworthiness decisions on loan requests.The company developed a PC-based expert system in 1992 to make preliminary creditworthiness decisions on loan requests. The company had experienced rapid, continuing growth and wanted the system to help ensure consistent, high-quality loan decisions.The company had experienced rapid, continuing growth and wanted the system to help ensure consistent, high-quality loan decisions. CLUES (Countrywide's Loan Underwriting Expert System) has about 400 rules.CLUES (Countrywide's Loan Underwriting Expert System) has about 400 rules. Examples of Expert Systems

65. 10.65 Countrywide tested the system by sending every loan application handled by a human underwriter to CLUES as well.Countrywide tested the system by sending every loan application handled by a human underwriter to CLUES as well. The system was refined until it agreed with the underwriters in 95 percent of the cases.The system was refined until it agreed with the underwriters in 95 percent of the cases. Countrywide will not rely on CLUES to reject loans, because the expert system cannot be programmed to handle exceptional situations such as those involving a self-employed person or complex financial schemes.Countrywide will not rely on CLUES to reject loans, because the expert system cannot be programmed to handle exceptional situations such as those involving a self-employed person or complex financial schemes. An underwriter will review all rejected loans and will make the final decision. CLUES has other benefits. Traditionally, an underwriter could handle six or seven applications a day.An underwriter will review all rejected loans and will make the final decision. CLUES has other benefits. Traditionally, an underwriter could handle six or seven applications a day. Using CLUES, the same underwriter can evaluate at least 16 per day.Using CLUES, the same underwriter can evaluate at least 16 per day. Examples of Expert Systems

66. 10.66 EXPERT SYSTEMS LIMITATIONS: EXPERT SYSTEMS LIMITATIONS: Often reduced to problems of classificationOften reduced to problems of classification Can be large, lengthy, expensiveCan be large, lengthy, expensive Maintaining knowledge base criticalMaintaining knowledge base critical Many managers unwilling to trust such systemsMany managers unwilling to trust such systems* AI

67. 10.67 CASE - BASED REASON (CBR) AI USES DATABASE OF CASES: USER DESCRIBES PROBLEMUSER DESCRIBES PROBLEM SYSTEM SEARCHES DATABASE FOR SIMILAR CASESSYSTEM SEARCHES DATABASE FOR SIMILAR CASES SYSTEM ASKS MORE QUESTIONSSYSTEM ASKS MORE QUESTIONS FINDS CLOSEST FITFINDS CLOSEST FIT MODIFIED AS REQUIREDMODIFIED AS REQUIRED* AI

68. 10.68

69. 10.69 Compaq Computer of Houston, Texas, gave purchasers of its Pagemarq printer case-based reasoning software to help reduce customer service costs.Compaq Computer of Houston, Texas, gave purchasers of its Pagemarq printer case-based reasoning software to help reduce customer service costs. Examples of CBR SYSTEMS Examples of CBR SYSTEMS

70. 10.70 c h a p t e r 12 MANAGING KNOWLEDGE: KNOWLEDGE WORK AND ARTIFICIAL INTELLIGENCE