Abstract

Computerizing the game of Bridge has not yet met with much success. The efforts to date have fallen short of any reasonable technical proficiency. The game does appear to be perfectly suited for an expert system, however, since the game can be segmented into three contexts (Bidding, Play of the Hand, and Defense), each context can be described by a set of rules, and a series of inferences can be used to fire those rules. Each of the contexts is reviewed, then Bidding is chosen for further research. This thesis claims that the set of all hands subdivides into 1 1 bidding classifications, based on a number of selection criteria. One of these subsets, Invitational Hands, is studied in detail Classic knowledge acquisition techniques are used to define Invitational Hands, assimilate the knowledge, then translate the facts, inferences, deductions and suppositions into a knowledge base. Changes in the state of the auction as bidding progresses are stored in state variables. These state variables are used to navigate the knowledge base to find the next bid. The interaction of state variable settings and facts firing rules in the knowledge base implement a frame architecture.

Library of Congress Subject Headings

Contract bridge--Data processing; Knowledge acquisition (Expert systems); Expert systems (Computer science)

Publication Date

1991

Document Type

Thesis

Department, Program, or Center

Computer Science (GCCIS)

Advisor

Anderson, Peter

Advisor/Committee Member

Lindenmayer, Raymond

Comments

Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works. Physical copy available through RIT's The Wallace Library at: GV1282.3 .S54 1991

Campus

RIT – Main Campus

Download

Share

COinS