Abstract:

The paper gives an example of the intelligent trade distinction prototype system according to the background of the knowledge for anti-money laundering. It combines with the feature extraction of semi-structured texts based on ontology and case-based reasoning. We process text mining through the suspi cious case reports submitted,and match with the cases which are in the case database. So, we can detect the highly suspicious transaction records. The case-based reasoning technology for screening large suspicious transactions is the successful expansion of the technical applications for ontology, and is the effective supplement for model-based reasoning. The description about domain knowledge makes it available to analyze the relationships between key
 words by ontology. The paper describes the method and the ideas of CBR, and illuminates it according to our example.

Key words: （ontology, case-based reasoning, feature extraction）