NigerScrab: A MACHINE-HUMAN BASED SYSTEM FOR CONTROL IMPLEMENTATIO IN COMPUTER BOARD GAME OF IMPERFECT INFORMATION

ABSTRACT
Computer games are important elements that have significantly approbated our culture, being a primary driving force for social and economic development. One of the most popular and most interesting games is Scrabble, a board game in which players compete by forming words on a common 15-by-15 grid. Our motivation for carrying out this research is attributed
to recent developments and studies in Artificial Intelligence (AI) games research, which have drifted from perfect information, whereby the players have full knowledge of game details, to imperfect information games, that is, games like Scrabble, where certain relevant details are withheld from the players. Hence, we take an in-depth look at how Artificial Intelligence is
applied in imperfect information game programming, being one of its most preferred uses.
The objective of the research was to identify some problems with existing methods and discover ways by which we can improve the game strength. Consequently, the proposed move generation algorithm of NigerScrab was able to do suffix search of 41,234 words using our model with cross check set as against 62,443 without cross check, thereby cutting off unnecessary word search. The system used a CPU time of 0.0022s as against 0.0034s without
cross check. Performance evaluation of NigerScrab revealed that the game strength has actually improved, having an overall average point per turn of 26.8 using a dictionary of approximately 80,000 words.
To achieve our objective, we formulated an improved model for evaluating the rack contents and generating moves using probability and heuristics search. Accordingly, we adopted the Object Oriented Analysis and Design Methodology for implementing the proposed model, using the Unified Modelling Language (UML) as the key to development; these could be viewed as part of our contributions to knowledge.