Title :
CTGE-ILP: Correlative Term Graph-Based Extended ILP
Author_Institution :
South China Normal Univ., Guangzhou
Abstract :
Now it is much more prevalent to mine complex data in the fields of data mining and the main methods include the experience and technology of no-structer data mining. On account of the flexible practicability and broad appliance of graph theory and technology, the research of graph-based data mining has presently been becoming an important direction in this domain. In this paper, knowledge discovery researching was studyed throgh the technology of correlative graph- based data mining. The concept of CTGE-ILP (Correlative Term Graph-based Extended Inductive Logic Programming) and the correlative term graph were bring forward, and then discussing the reasoning technology by combining correlative term graph theory and inductive logic programming technic. Next we design the technology of refutable consequence in CTGE-ILP and the corresponding algorithm. At last an prototype integrating both the idea and technic above was opened up and turned out its validity via the data demonstration.
Keywords :
data mining; graph theory; inductive logic programming; inference mechanisms; correlative graph-based data mining; correlative term graph theory; extended inductive logic programming; knowledge discovery; no-structer data mining; reasoning technology; Algorithm design and analysis; Data mining; Educational institutions; Educational technology; Graph theory; Home appliances; Logic programming; Machine learning; Pattern recognition; Prototypes;
Conference_Titel :
Semantics, Knowledge and Grid, Third International Conference on
Conference_Location :
Shan Xi
Print_ISBN :
0-7695-3007-9
Electronic_ISBN :
978-0-7695-3007-9
DOI :
10.1109/SKG.2007.93