Introduction to Discovering Conditional Functional Dependencies Java Project:
The main objective of this paper is to give idea of conditional functional dependencies. As usage of computers is increasing are day to day, the storage of data also is more. The data has to be cleaned an removed if it is not useful.
Data cleaning should be done automatically by finding out the errors. Hence this gave rise to conditional functional dependencies. CFDs are considered as a extension of functional dependencies. CFD supports patterns of related constants, and hence it can be used as rules for cleaning relational data.
For discovering CFD three methods are used . First, CFD Miner is used to discover CFDs whish have constant patterns. These constant CFDs are important for identifying an object to clean .The two algorithms which are used for discovering general CFDs is CTANE and FastCFD, CTANE is considered as extension of TANE which is used in mining of FDs.
FastCFD is based on the depth-first approach and it reduces the search space. When the relation is large, CTANE is said to work well. FastCFD works better than CTANE in terms of arity, when the arity of the relation is large. Using CFDMiner is faster than CTANE and FastCFD. These algorithms used for cleaning are based on type of application by the user.
Download Discovering Conditional Functional Dependencies Java Project.