Scoring Engine

SEON provides a completely transparent decision and score as a response. SEON’s algorithm is based on industry and geography-specific pre-added rules, custom rules by the client, and machine-learning rule suggestions.
 

Default Rules

Pre-added rules tailored to industry and geographic area. It’s possible to turn them on/off and modify the score, but you can’t add new ones or delete them. 

 

Machine Learning Rules

Automatically generated rules based on your feedback. It’s possible to turn them on/off, but you can’t add new ones or delete them.

 

Rule Testing

Simulate rules and test their efficiency using your historical data. Effortlessly visualize what effects the proposed rule would have on your transactions via a confusion matrix and see the exact number and ratio of transactions that would fall in DECLINE or APPROVE states and their monetary value.

Rule testing in action

We ensure a fully customizable service. In order to optimize the scoring engine, admins can create rules based on any relevant logic. On the rules page, all of the existing rules are listed with the option to modify or delete them.

 

Rules and Rule Parameters

Rules may contain multiple parameters separated by specific operators. These rules represent the various branches in our decision tree. Meanwhile, rule parameters are the nodes.

 

Custom Rules

Add your custom rules to the algorithm and always stay in control of the decision model. We work with any kind of custom rules and the Scoring Engine provides a flexible environment for fine-tuning. 

 

Scoring Settings

With each rule, you can add, extract a specific score from a transaction final score or you can automatically put the transaction in a specific state.

Rules may have three types of effects. A rule can either change the score of a transaction, it can set the state of a transaction or it can black/whitelist a certain field. Rules are triggered according to rule hierarchy.

 

Rule Parameters

It is possible to set up rule parameters based on four different types of logical inferences. We call these rule parameter types.

Multiple rule parameters can be added to a rule by clicking the arrow below the parameter preview.

The added parameters can be modified and deleted in the list below.