Tiered ML Dispute Resolution for Accurate Marketplace Decisions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing back-end computing systems in online marketplaces often fail to properly or incorrectly resolve disputes between buyers and sellers, leading to further issues.
Innovation Solution
A tiered machine learning engine is employed to analyze dispute resolution information, utilizing multiple machine learning modules to identify dispute reasons and determine resolutions, allowing for continuous updates of specific tiers without requiring a full system overhaul.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a conventional back-end computing system is used to resolve disputes, then the system structure is simple, but the dispute resolution accuracy is low and errors occur
Solution Approach 1:
The machine learning engine is divided into multiple tiers, with each tier containing specialized machine learning modules that analyze different aspects of dispute information. This segmentation allows each module to focus on specific patterns and improve overall resolution accuracy while maintaining manageable complexity through modular design.
Solution Approach 2:
The patent introduces a sophisticated machine learning engine as an intermediary between the buyer and seller dispute inputs and the final resolution. This intermediary processes and analyzes the dispute information through multiple tiers of analysis, transforming raw data into accurate resolution decisions.
2Productivity
If the entire machine learning engine is updated simultaneously, then the model remains consistent, but the computing resources and time required increase significantly
Solution Approach 1:
The machine learning engine is segmented into multiple independent tiers that can be updated separately. This allows the system to update only the specific tier that needs improvement rather than the entire engine, significantly reducing update time and computational resources while maintaining consistency within each tier.
Solution Approach 2:
The system allows dynamic updating of individual tiers without requiring a complete system overhaul. Each tier can be independently trained, validated, and deployed, enabling continuous improvement of dispute resolution capabilities over time without system-wide disruptions.
3Productivity
If a single-tier machine learning engine is used, then the system is simpler to manage, but it cannot handle complex disputes at scale efficiently
Solution Approach 1:
The multi-tier architecture segments the dispute analysis process into specialized levels, with each tier handling specific aspects of dispute evaluation. This segmentation enables the system to process complex disputes efficiently by distributing the analytical workload across multiple specialized modules rather than overwhelming a single-tier system.
Solution Approach 2:
The patent adds dimensional complexity to the machine learning engine by introducing multiple tiers of analysis. This dimensional expansion allows the system to handle increasingly complex disputes by adding analytical depth through additional tiers, scaling the system's capability to match dispute complexity.
Data Source
AI summary
Systems and methods including one or more processors and one or more non-transitory computer readable media storing computing instructions that, when executed on the one or more processors, perform: receiving dispute resolution information from a user corresponding to a dispute, wherein the dispute resolution information comprises a first data input and a second data input; analyzing, with a first machine learning module in a first tier of a machine learning engine, the dispute resolution information from the user to identify a dispute reason; analyzing, with at least one of second and third machine learning modules in a second tier of the machine learning engine, the dispute resolution information based on the dispute reason, as identified; and determining a dispute resolution based on an output of the second tier of the machine learning engine to resolve the dispute. Other embodiments are disclosed herein.