Research And Implementation Of Search Algorithms Of Prolog Fact-database On Raw Device

In today's world large numbers of data are produced every day in all walks of life. However, the primary concern of users is how to find the exact information from among the vast data in an efficient way and soonest possible. That is to say, the requirement of users for searching the right information becomes increasingly higher than before. The index mechanism effectively speeds up the response given by the system.Prolog, as one the most influential AI languages, has won the attention and has been widely used as well in the information domain as a result of Prolog's remarkable advantages in artificial intelligentization. This essay is based on the study of the search algorithm of Prolog fact base of raw device, a special device to read computer character without format or the buffering of the file system.The completion of the index mechanism in Prolog system by using the multi-level indexing technology of the B+ tree of raw device enables its user to quickly find the right information without the support of other software, which not only omits the intermediate process of data exchanges between the program and other software, but facilitates maintenance of this system.Thanks to the B+ tree index technology, the B+tree is finally set up by using the bottom-up methodology. It enhances the speed of search result and formulates the search module imbedded in Prolog. The emphasis of this essay is laid on the design of B+tree's data structure, the composition of key words, and the construction and searching of B+tree.The essay begins with the background of raw device study and knowledge of Prolog. After that, the essay briefs on related concepts and contents of current index system. In the end the essay propounds on the index algorithm.