Research On Learning-based Method For Solution Of Nine Puzzle Problem And Its Application

The research on Computer Game has never stopped since artificial intelligence came into being as a discipline from 1950s. It plays an important part in AI. In which, many universities and institutes have made great achievements. Compared with the game in man-machine chess, the research on computer game in games like Nine Puzzle, cards, magic cube has been less. Recently,the traditional way at present to solve the mathematical intellectual problems are low efficient.Whether the system has the ability to learn is an important sign of its "intelligence". Based on the traditional ways, a learning-based method to solve Nine Puzzle problem is thus proposed. This method simulates the process of human learning. Though each game, it can learn directly form the training examples, each state of the chessboard and correct steps, experience and knowledge, can be stored in "brain" permanently. As the experience is accumulated, its ability to solve the problem is also improved.Automatic Chinese segmentation is a key task in natural language processing (NLF) and computational linguistics. Its research and application are of great theoretical and practical significance. The learning-based method proposed in this paper is applied in Chinese text-segmentation. The software-system of Chinese text-segmentation can autonomously find many new phrases and learn them.The method researched in this paper can also be used to solve some other problems related to fixed resources distribution, for example, military expert system, military roberts, achieves management system, equipment automatic maintenance system, intelligent testing system, large scale data searching system, etc.