Week 7 Case-based Reasoning System CBR 1

Week 7 Case-based Reasoning System (CBR) 1

Case Scenario ABC Enterprise, a provider of multimedia content services always receives many incoming calls from subscribers asking anything about their services. For this reason, the company is considering to setup large customer service department to handle varied customer questions. The CEO of the company, Mr. Ridzuan is aware that the creation of such department will incur huge expenditure to maintain their services. Therefore, he wishes that there is an online system to handle specific subscriber problems regarding their content services. The system must be able to answer simple query problems and after searching using current knowledge base of content services, will try to display or suggest customer similar solutions already solved from previous related problems. 2

Why CBR ? n Problem: Most of the time the trouble in building Expert Systems comes from trying to fit experience into rules. Experience RULES 3

Why CBR? Extremely effective in complex cases. n Justification - Human thinking does not use logic (or reasoning from first principle) n CBR is the essence of how people reason from experience. n Process the right information retrieved at the right time. n 4

What is a Case? Defines a problem in natural language descriptions and answers to questions, and associates with each situation a proper business action. n Experience-based records. n 5

Case-based Reasoning (CBR) n n CBR method => adapt solutions used to solve old problems for new problems. CBR: n n Finds cases that solved problems similar to the current one, and Adapts the previous solution or solutions to fit the current problem, while considering any difference between the two situations. 6

Finding Relevant Cases Involves: Characterizing the input problem, by assigning appropriate features to it. n Retrieving the cases with those features. n Picking the case(s) that best match the input best. n 7

Case Representation n Example: key features of a car Year Model Make Options Condition Mileage 8

The CBR Process 1. 2. Find relevant cases in storage that have solved problems similar to the current problem (use similarity metrics). Adapt the previous solutions to fit the current problem context. Similarity Metrics (Distance Metrics) Solution Adaptation Methods Case Retrieval Case Adaptation Historic Case Library Test Solution New Problem Input (Source: Adapted from Marakas) Conclusion 9

Example: (Nearest Neighbors) test_scores (100%) age (years) name weight (kg) height (cm) David 108. 4 177 78 31 Daniel 88. 2 183 60 25 Patrick 81 175 54 27 John 104 198 55 38 Suppose you have a database of people. David, Daniel, Patrick, and John are people's names all stored in this database as instances. Some attributes from the database could be < weight, height, test_scores, and age >, so instances are the records (rows) of table people and the attributes are columns. 10

Cont. n Now, we have an Euclidean geometry of 4 -dimensional space (n=4), because we have four attributes < weight, height, test_scores, and age >; note that attribute name is omitted here, because it is symbolic (non-numeric). We can find the Euclidean distance between the instance David and instance John by using Equation 1. The coordinates in our 4 D space for David = (108. 4, 177, 78, 31) and for John = (104, 198, 55, 38) with the calculation shown below: d(David, John) = = = 32. 2236 (round to 4 decimal places) 11

CBR Example: PC Shopping n You (the buyer) will specify your target preference (query point xq ) for the computer that is closest or similar to the attributes you want. The online store has a database with different types of computers (Training examples). The target query (specification for a computer to buy) will be ranked according to its nearest neighbours. 12

CBR Selection Engine Result 13

Example of a successful system n n CBR is particularly used for help-desk applications. For instance the COMPAQ SMART system. 14

Example of a successful system (COMPAQ SMART system) n The problem was that: ¨ Thousands of customers were calling Compaq directly every day, requesting support. ¨ Many of the staff were new; there was a major training problem. ¨ There was a need for consistent & accurate answers and responses ¨ There was a need for retention of corporate knowledge. 15

Example of a successful system n n The COMPAQ SMART system, once developed and installed, succeeded in solving 85 -95% of calls. Typical time to solve a problem was less than 2 minutes. 16

Advantages of CBR n Case-based reasoning: ¨ tends to focus on the problem's essential features. ¨ can solve problems in domains that are only partially understood. ¨ can provide solutions when no algorithmic method is available. ¨ can interpret open-ended and ill-defined concepts. 17

When to Use CBR? n n n . . . users have to be supported and advised. . . cases can easily be identified and created, products are simple to describe. . . easy maintenance of the case based is desired. . . no 100% coverage of the domain is required. . . Similarity based retrieval is acceptable fast CBR technology can be understood as the fusion of these concepts whereby the advantages of knowledge-based systems are linked to existing data. Database Expert System CBR Information Retrieval 18

CBR Tools - Examples n n n ART*Enterprise and CBR Express (Inference Corporation) KATE (Acknosoft) Re. Mind (Cognitive Systems Inc. ) Support for information security products (Symantec): http: //www. symantec. com/search/ 19

Intelligence Density Dimension n n Accuracy Response time Flexibility Independence from experts 20