Research On The Theory And Application Of Incomplete Information Database

Incomplete information database and SQL query processing in relational database are two important fields in database theory. They do not only have important theoretical value, but also have great practical value. The medium logic predicate calculus system MFM is a new logic system modified from the existing medium logic system. It is fit for dealing with incomplete information theoretically and practically. Firstly, to deal with the queries processing in the incomplete information perfectly, the SQL queries are modified to reslove the incomplete information advisably. Secondly, according to the case that some tuples are missing or the databases are distributed, we give the definition of partitial databsase, and put forward to the corresponding solutions and the relevant theorems and proofs in order to obtain complete answers. Then we discuss relation between answer-completeness, answer-correctness and data update,which lay the foundation for incomplete queries in theory. Thirdly, we consider rough sets theory based on null values, and propose a new method to classify attributes in the rough sets with three different null values, also give the equivalence and tolerance relations between objects, define three characteristic sets according to the different needs of decision maker, which have important guidance meaning in decision support system, data mining techniques. At last, with the De Morgan law, we define a new negation connective, construct a new four-valued logic system, and give the comparison operation and selection operation.