Online Payment Risk Assessment via Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Online payment services face significant risks from fraudulent transactions, including collusion between vendors and consumers, leading to financial losses due to chargebacks, as existing fraud detection systems are inadequate in preventing such fraudulent activities.
Innovation Solution
A system that uses a machine learning model to assess risks by aggregating historical transactions, personal attributes, and credit information of both vendors and consumers, generating a risk score to determine whether to advance funds for a transaction, and taking actions such as declining transactions or reporting suspicious accounts to prevent fraud.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the online service processes all transactions without risk assessment, then transaction volume and productivity are maintained, but financial losses from fraudulent chargebacks increase significantly
Solution Approach 1:
The system performs risk assessment and generates risk scores for transactions before funds are advanced to vendors. This preliminary evaluation allows the online service to identify and flag potentially fraudulent transactions in advance, preventing financial losses while maintaining processing efficiency for legitimate transactions
Solution Approach 2:
The risk score acts as an intermediary mechanism between transaction processing and fund advancement. By introducing this intermediate evaluation layer, the system can selectively advance funds based on risk levels, thereby reducing chargeback losses without significantly impacting overall transaction volume
2Loss of energy
If the online service implements comprehensive risk assessment for all transactions, then financial losses from fraud are minimized, but system complexity and processing time increase
Solution Approach 1:
The risk assessment system operates autonomously using machine learning models that automatically evaluate transactions, generate risk scores, and determine fund advancement decisions without requiring manual intervention. This self-service approach minimizes system complexity while maintaining comprehensive fraud protection
Solution Approach 2:
The system evaluates multiple parameters including vendor history, consumer credit information, and transaction characteristics to generate a composite risk score. By transforming diverse data parameters into a single risk metric, the system achieves comprehensive assessment without proportionally increasing complexity
3Ease of operation
If funds are advanced to all vendors for all transactions, then vendor satisfaction and service quality are maintained, but exposure to financial risk from fraudulent transactions increases
Solution Approach 1:
The system applies different fund advancement decisions to different transactions based on their specific risk profiles. Low-risk transactions receive prompt fund advancement to maintain vendor satisfaction, while high-risk transactions have fund advancement delayed or denied, thereby protecting against fraud while preserving service quality for legitimate vendors
Solution Approach 2:
The system advances funds partially and selectively rather than universally. By advancing funds only for transactions below certain risk thresholds, the system maintains vendor satisfaction for legitimate transactions while limiting exposure to fraudulent ones, achieving a balanced approach between service quality and risk management
Data Source
AI summary
A system receives a request for payment of a transaction between a vendor and a consumer, and sends a first request to a database associated with the online service for historical transactions and personal attributes of the vendor concurrently with sending a second request to a number of third-party services for credit information and personal attributes of the consumer. The system receives information responsive to the first and second requests from the database and the third-party services, respectively, and obtains a risk score for the transaction based on an application of one or more risk assessment rules to the received information by a machine learning model trained with at least the historical transactions and the personal attributes of the vendor. In some aspects, the system determines whether to advance funds to the vendor, prior to requesting payment from a consumer account, based at least in part on the risk score.