Abstract:

Generally, there are consanguineous relationships between financial crimes and the flow oi iunds, especially m money-laundering. Instead of taking a m  ore traditional approach such as starting with an inspection of the data, formulating questions or hypotheses interesting to us and then devising an ana lysis approach to answer them, We try a new different route： we make use of constraint-based unsupervised link discovery to find the user-interested in nformation, because traditional approaches have difficulty in detecting unknown or unfamiliar knowledge patterns. This method can be described as follow  s： we judge if the target node accords with the constraints, and then analyze foreign exchange flows by computing the ＂interestingness＂ and ＂contrib bution＂ of paths. Finally, we can acquire interesting information, supply clues and make decisions in detecting foreign financial crimes.

Key words: （link discovery, constraint-based unsupervised link discovery, interestingness, contribution, money-laundering）