Automated Item Network for Competing Product Discovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The manual identification and monitoring of competing data records in information processing systems are resource-intensive and often performed infrequently, leading to delays in responding to market changes, as they rely on subjective domain expert knowledge and are prone to bias.
Innovation Solution
A method that generates an item network by connecting products based on type, price ratio, and pairwise configuration similarity scores, using collaborative filtering and network analysis to automatically identify competing products, allowing for frequent updates and data-driven insights.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual identification and monitoring of related data records is performed, then domain expert knowledge can be applied, but the process becomes resource-intensive and time-consuming
Solution Approach 1:
The patent replaces manual mechanical processes with automated computational systems. Specifically, it substitutes human expert manual review with machine learning models (collaborative filtering algorithms) that automatically identify related data records, and replaces manual network construction with automated graph generation based on learned relationships, thereby eliminating resource constraints while maintaining identification quality
Solution Approach 2:
The system enables self-service by allowing the automated identification system to continuously monitor and update related data records without human intervention. The collaborative filtering model automatically learns from data patterns and updates relationships, and the system self-updates the item network structure based on new data, eliminating the need for periodic manual execution
2Loss of energy
If manual monitoring is performed occasionally (monthly or quarterly), then resources are conserved, but delays occur in responding to market changes
Solution Approach 1:
The patent implements continuous automated monitoring that operates without interruption between data updates. The system continuously maintains the item network structure and automatically identifies related records as new data arrives, ensuring real-time responsiveness to market changes while eliminating the periodic gaps inherent in manual monthly or quarterly review processes
Solution Approach 2:
The system dynamically adapts to changing market conditions by continuously updating the collaborative filtering model with new data and automatically re-computing item relationships. The item network structure dynamically evolves as new data records are added, allowing the system to respond flexibly to market changes without requiring manual reconfiguration or fixed scheduling intervals
3Adaptability or versatility
If manual identification is used, then subjective domain expert knowledge is applied, but bias and inconsistency are introduced
Solution Approach 1:
The patent transforms subjective domain expert knowledge into objective computational parameters by encoding domain expertise into the collaborative filtering algorithm's feature selection and weighting mechanisms. The system maintains adaptability to domain-specific patterns while ensuring consistent, reproducible results through parameterized model execution that eliminates human bias and variability in identification processes
Data Source
AI summary
Techniques are provided for automatic discovery of data records. One method comprises obtaining data records each corresponding to a different item and comprising features extracted from a data source, wherein the data records identify related items identified using a collaborative filter that relates items based on user preferences; generating an item network comprising multiple nodes each corresponding to a different item, where two nodes are connected by an edge based on: (i) an item type of the two nodes, (ii) a ratio of numerical values associated with the two nodes, and/or (iii) a pairwise configuration similarity score for the two nodes; clustering the nodes into node clusters based on topological properties of the item network; and identifying items related to a given item that (i) share an edge with the given item and (ii) are in a node cluster comprising a node of the given item.