tuzijun111 / Schema-Matching

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

DySM

There are four algorithms for incremental schema matching. And they can be run by the main functions in the Experiment, Cluster and DSframe classes.

In the Experiment class, we first integrate sources from source1 to source 8 one by one and we take the integrating results as the initial clustering result. (This is the first algorithm, and we can call it as naive schema matching algorithm). Then we apply the greedy DBindex algorithm (DBGreedy, the second algorithm) on the initial clustering result. The two algorithms can be run in the main function of the Experiment class.

In the Cluster class, the main function contains our proposed CenterCluster algorithm (the third algorithm). This algorithm is desinged as an incremental algorithm. So the batch algorithm is the same with its incremental algorithm.

In the DSframe Class, the main function contains the incremental algorithm for DBindex clustering (it is called IncrementalDB, the forth algorithm).

About


Languages

Language:Java 100.0%