INCOP System: Intelligent Connection Profiling System

 

INCOP System: Intelligent Connection Profiling System

 

Fraud introduction:

Adakami divides fraud into bad intermediary fraud, negative information on the Internet, fraudulent use of others to apply for loans, data fraud, and fraud gangs.

1. Fraudulent use of others to apply for loans: Fraudsters illegally obtain other people's account numbers, passwords, ID numbers, etc., pretending to be victims to apply for loans, and profit from fraud.

2. Data fraud: Fraudsters maliciously forge false data and use these data to commit fraud.

3. Fraudulent group: a group of unscrupulous users gathered for the purpose of fraud, and fraudulent behavior in the form of a group to cheat to obtain credit.

4. Bad intermediary fraud: Some people in the gang do not use fraud as a means of profit, but a group of criminal gangs that help others to defraud and charge high handling fees as their profit method. Compared with the previous three, bad intermediary fraud is more harmful, and the entire organization has strict discipline, better concealment, and stronger professionalism.

5. There is negative information on the Internet: customers who can fake or forge information on the Internet, collect customer information that is overdue all the year round, and cooperate with other third parties to enrich the blacklist database.

Among them, bad intermediary fraud and fraud gangs are the most serious and hidden frauds, and it is difficult for traditional investigations to accurately detect fraudulent customers. In order to reduce the risk of fraud, adakami has established a fraud manual investigation team and developed the  INCOP anti-fraud system.

 

Professional anti-fraud investigation team:

Professional anti-fraud personnel can effectively identify fraud .They use the system to provide as comprehensive information as possible, such as customer equipment information, location information, activity record information, etc. to conduct information investigations, find fraudulent customers and add them to the blacklist. Adakami has conducted in-depth research on anti-fraud and formed a unique anti-fraud investigation system. It guides the investigation direction through associations and fraud chains, uses business experience to find high-risk suspects, and uses data in the system for auxiliary verification. , And finally investigate and verify, which greatly reduces the fraud risk of adakami as a whole.

With the continuous rapid growth of business volume, we have not only accumulated a lot of experience, but also obtained a large number of fraudulent user samples. At this time, combined with adakami's AI capabilities and engineering capabilities, the  INCOP anti-fraud system came into being.

 

Adakami's high-tech background:

Relying on a strong scientific and technological team, our big data and artificial intelligence technology have fully covered the three links in the lending business process: pre-loan, in-loan, and post-loan: accurate customer acquisition systems before lending; intelligence in lending Risk control system,  INCOP anti-fraud; post-loan model system and intelligent collection after loan; in addition, it also includes intelligent customer service. The application of these new technologies has greatly improved our operational efficiency.

 

INCOP introduction:

INCOP relies on the relationship between different customers, by restoring the social network relationship between users, building a complex network composed of tens of millions of nodes, such as whether customers use the same mobile phone to apply for loans, whether to use the same geographic location, etc. Association relationship to discover different roles and relationships in group fraud, distinguish fraudsters, suspicious users and normal customers, and effectively identify and prevent fraud by effectively identifying abnormal fraudsters.

 

Data advantages of INCOP:

Adakami relies on the rich anti-fraud business experience of 500 million linked data, 6 million users, 9 million pattern samples, and 400,000 blacklist data, builds a powerful model system, and independently develops risk identification for the entire process of anti-fraud investigations Sharp weapon-" INCOP". Compared with traditional anti-fraud, " INCOP" has a visual relationship network that can be expanded up to 6 layers, which can help the anti-fraud team accurately find the relationship, and can achieve linkage with the business back-end, thereby helping to greatly improve the effectiveness of anti-fraud investigations . After " INCOP" was put into use, the investigation efficiency of Adakami's anti-fraud team has increased”

 

The birth and function of INCOP:

The original intention and core appeal of INCOP is to improve the effectiveness of anti-fraud investigations. All functions of the system are developed around this core appeal.

 

1. Use of Knowledge Graph

We have introduced a knowledge graph, which uses the associations between all users and users to form a graph, takes all customers as points on the graph, and all associations as lines, and uses different icons to identify them, thus constructing an association network. Intuitively display the associated network of fraudulent users to investigators, so that it is convenient to find associated fraud gangs at one time.

 

2. Judgment of suspicious behavior

Summarize the characteristics of the abnormal customer through the experience of the investigator and extract them into useful rules, and add them to the  INCOP system. During the investigation, the investigation can be conducted first, and different investigation methods can be used according to the different characteristics of the customer.

 

3. AI image recognition function

In Der Spiegel, we have also introduced more AI components to help improve the efficiency of investigators, such as image recognition. By showing customers who are suspected of using the same face to anti-fraud personnel, abnormal customers can be accurately located.

 

 

4. Find the key person in the gang

It is difficult for users' social relationships to be directly related to fraud. However, bad intermediaries are the central figure in group fraud and are directly related to fraud. Through the rule touch function on Der Spiegel, it can help anti-fraud employees identify key fraudsters more quickly, thereby eliminating all associated customers.

 

From the beginning of its launch, Adakami has insisted on a pure credit unsecured loan method, firmly believes in the power of technological risk control, and tries to protect the security of lenders' funds to the greatest extent possible. We will never stop on the road of using technology to drive risk control efficiency.