Network Traffic Blocking via Reseller Interaction Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
First parties struggle to effectively identify reseller users in network interactions, which can negatively impact their ability to offer products to individual users, leading to potential losses in sales.
Innovation Solution
The system uses data associated with reseller users' interactions with virtual locations to generate scores indicative of their likelihood of being resellers, allowing first parties to identify and potentially block such users.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If first parties do not implement identification systems for reseller users, then network interactions remain simple and unmonitored, but they cannot identify and limit reseller users leading to sales losses
Solution Approach 1:
The system performs preliminary actions by collecting and storing interaction data before final identification occurs. Historical interaction data is accumulated and analyzed to build profiles that enable future reseller detection, allowing the system to prepare identification capabilities in advance rather than reacting to each interaction individually
Solution Approach 2:
The system introduces an intermediary analysis layer between raw interaction data and final identification decisions. This intermediary component processes and scores interactions based on predefined criteria, acting as a mediator that transforms raw data into actionable identification results without requiring direct complex analysis at the point of interaction
2Reliability
If first parties block all network interactions to protect sales, then individual user sales are protected, but legitimate users are also blocked causing loss of business
Solution Approach 1:
The system applies local quality by evaluating each interaction and user profile with differentiated criteria. Instead of uniform blocking, it assigns different weights and thresholds to different interaction patterns, allowing legitimate users with normal behavior patterns to pass through while reseller patterns are flagged. Each user's history and context is considered individually
Solution Approach 2:
The system changes parameters dynamically by adjusting identification thresholds and scoring criteria based on accumulated data. The scoring model evolves by learning from new interaction patterns, allowing the system to adapt its sensitivity to reseller detection over time while maintaining stability for legitimate users through continuous parameter optimization
3Measurement precision
If first parties implement comprehensive monitoring of all network interactions, then reseller identification improves, but processing time and computational resources increase
Solution Approach 1:
The system segments the interaction analysis process into distinct components: data collection, pattern matching, scoring, and decision-making. By dividing the complex monitoring task into separate modular functions, each component can be optimized independently and processed more efficiently, reducing overall processing time while maintaining precision
Solution Approach 2:
The system applies partial action by focusing analysis only on interactions that meet certain preliminary criteria or show suspicious patterns. Not all interactions require full-depth analysis; the system performs targeted examination only where needed, reducing computational overhead while maintaining high detection precision for actual reseller behavior
Data Source
AI summary
Systems and methods are provided for identifying and blocking network interactions. One example computer-implemented method includes accessing, by a server computing device, access data associated with an interaction between a user and a virtual location of a first party, where the access data includes a common data element specific to the interaction and at least one click behavior, and accessing identity data associated with the interaction between the user and the virtual location, where the identity data includes the common data element specific to the interaction and identifying data associated with the user. The method then includes joining the access data and the identity data based on the common data element, generating a score based on the joined data and a model, where the sore is indicative of a probability of a type of the user, and reporting the generated score to the first party.