Detecting the financial offences
Financial offences are getting more and more difficult to detect. Fraudsters are making up:
Standard mechanisms for fraud tracking (also those, based on Machine Learning), to a significant extent are not able to discover the situation mentioned above. It’s so, because analysed are the data, not the connections. If we connect persons, that use the same ID card no. or personal identifcation no., or persons hsharing the same phone no. or email address, or persons sharing the same living address etc., it will become obvious that these persons will be identified as “mutually connected” – although in conventional databases they will remain strangers. Graph database in-built algorithms allow for discovering connections and the groups of information, which in turn helps in finding the fraud chains.
The structure of GraphIQ tool enables to adapt the data structures to ever-changing requirements (e.g. fraud detection tactics).
Defining adequate conditions, e.g.:
we can draw a clearly designed graph, containing nodes and relations within previously indicated transactions and conditions, and subject it to further analyses. It is also possible to make use of external data.
There are also algorithms, predicting a specific behaviour of specific persons, based on their “friends” (persons directly related to each other). Defining the connections between different persons may be of a crucial importance, now and in the future.