Research And Implementation About An Incremental Knowledge Acquisition Algorithms Based On Rough Set

The present database system may realize functions highly effective, such as data input, inquiry, statistics, but is unable to find the relations and the rules in the data, will be unable to forecast the future according to the existing data, and lack of data mining hidden knowledge means, led to the "data explosion but knowledge poverty". With the development of database and artificial intelligence, Knowledge Discovery in Data (KDD) has become a new database technology in the past few years. With KDD technology, we can turn information into knowledge and found the "knowledge nuggets" from the"data mining". Wherefore, Research efficient intelligent knowledge discovery methods have great practical significance.Rough set theory that was put forward by Pole Z.Pawlak in 1982 is a new data analysis theory of analyzing and dealing with uncertain and incomplete data. It makes use of the equivalence relations to measure the indetermination degree of knowledge and it doesn't need any knowledge outside of the data which needs to be processed. Therefore the error caused by subjective appraisal can be avoided. With 20 years development, the rough set theory not only in the theoretical study of continuously improve itself, but also in other fields have also been successful in applications, such as machine learning, decision analysis, approximate reasoning, image processing, expert systems, process control, conflict analysis, knowledge discovery in databases, medical diagnosis, such as financial data analysis has received more successful applications.In recent years, research in the rough set theory for solving the minimum attribute reduction and a smaller reduction, as well as getting a simplest rule set algorithm has been going on for some studies, but these studies are directed at the static data. The database is dynamic. Many researchers suggested that the database knowledge discovery algorithm is the incremental. Attribute minimum incremental reduction algorithm and incremental updating the concept lattice algorithm has begun to be studied, but the research in incremental rule extract algorithm is still relatively small. On the basis of the above work, this paper studies the incremental rule extract algorithm based on rough set.In this paper, the main work done: Introduce rough set theory course of the development, features, applications and basic theory; Put forward the definitions of"relevant"and"independent"between the objects in the decision table; Which rules need to be updated when a new object is put into the decision table; proposed an incremental rule extraction algorithm based on rough set and search tree. The depth-first search algorithm takes the extracted rules as heuristics information and greatly improve the efficiency of the algorithm; the algorithm has better space complexity with the same time complexity, which not to establish the distinguish matrix in the traditional algorithm. Because the reality of the decision table is often inconsistent, the paper provide the algorithm based on variable precision rough set in order to extract the uncertainty rules. Finally, by contrast experiment proved the correctness and effectiveness.The main innovation lies:(1) Put forward an incremental rule extract algorithm based on rough set and search tree. The algorithm is a depth-first search algorithm and takes the extracted rules as heuristics information, greatly improve the efficiency of the algorithm.(2) Provide the algorithm based on variable precision rough set, which can extracts the uncertainty rules efficiently according to the rule confidence.