Font Size: a A A

A Fuzzy Decision Tree Model And Its Applications

Posted on:2007-09-13Degree:MasterType:Thesis
Country:ChinaCandidate:D Z DingFull Text:PDF
GTID:2178360212957239Subject:Software engineering
Abstract/Summary:PDF Full Text Request
Intelligent call center that call center integrated with artificial intelligence especially expert system, makes use of expert system to provide a high-level service for call center users, which improves the service level of call center and lowers work intension. For the extensive applications of general telephones, visiting expert system with general telephones is a wonderful method. And that many questions like the inconvenience of the interaction between human and machine, effective reasoning, etc. need to be resolved. To reduce interaction times between human and machine, an effective reasoning algorithm must be provided .The reasoning based on fuzzy decision tree (FDT) is one of some available methods. Experts resolve domain questions through fuzzy acknowledge generally. Thus, the essential task of this research is to translate fuzzy production rules into FDT reasoning mode and to build the reasoning mechanism for a call center.(1) This article puts forward two algorithms to translate the knowledge of fuzzy rules into FDT in a Call Center System. The core concept of the algorithms is the classification ambiguity called G. (2) This paper designs fuzzy knowledge expression methods based on FDT, a certainty factor model, and a heuristic search technology.Based on the acknowledging of the classes of uncertainties, and after making a lot of testifying calculations, we put aside the fuzzy entropy which is the core concept in FID3 algorithm and is familiar to us, and we adopt the classification ambiguity which is used in Min-Ambiguity algorithm as the essential concept to built the fuzzy reasoning model based on FDT. At the phrase of constructing the model, we divide the circumstances into 3 types & we discuss it with the calculating examples: The first is with no realm expert knowledge but a great deal of case data; the second, with the realm knowledge, also a great deal of case data and so on. At the fuzzy reasoning stage, we firstly propose a heuristic search algorithm based on the model, and then we compare the reasoning application of call center to the common reasoning applications. At the end, we make algorithm designs of the practical project.The design and implementation of an intelligent call center is introduced in the end of this paper. It is mainly about the fuzzy expert system design, sub-system design etc. Used the weather and sport training set and rules, an intelligent call center model with a fuzzy export sub-system module is realized.
Keywords/Search Tags:Fuzzy decision tree (FDT), Heuristic Search, Call Center, Fuzzy Production Rule, Fuzzy Reasoning
PDF Full Text Request
Related items