Decision Tree Matching for Search Keywords and Product Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining the matching degree between search keywords and product information are resource-intensive, prone to low accuracy, and costly due to the need for extensive human labeling and maintenance, especially when system changes occur.
Innovation Solution
A method and apparatus using a decision tree to match and determine features between search keywords and product information, eliminating the need for linear models and human-labeled text correlation features, thereby simplifying the process and reducing maintenance costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If human experience labeling is used to label text correlation feature values for search keyword and product information pairs, then the matching degree can be determined, but the system consumes excessive resources and incurs high maintenance costs due to the large number of pairs requiring labeling
Solution Approach 1:
The system uses automated algorithms to determine text correlation feature values without requiring human annotators. The model independently processes search keywords and product information to generate matching scores, eliminating the need for manual labeling of all search pairs while maintaining determination accuracy.
2Measurement precision
If human experience labeling is used to label text correlation feature values, then the matching degree can be determined, but the accuracy of labeling is very low
Solution Approach 1:
The patent replaces the mechanical human labeling process with an automated computational model. Instead of relying on human annotators to assign text correlation feature values, the system uses algorithmic processing to automatically determine these values, thereby eliminating human error and improving labeling accuracy.
3Adaptability or versatility
If the number of text correlation features is increased or decreased and the system is upgraded, then the system can adapt to new requirements, but the values of the text correlation features need to be relabeled, resulting in high maintenance cost
Solution Approach 1:
The system is designed to dynamically adapt to changes in the number and types of text correlation features without requiring complete relabeling. When features are added or removed, the automated model can recalculate relevant values based on the updated feature set, maintaining system adaptability while avoiding the high costs associated with manual relabeling of all search pairs.
Data Source
AI summary
A method and an apparatus of matching an object to be displayed are disclosed. The method includes obtaining a plurality of search keywords and released product information and grouping each of the plurality of search keywords with the released product information to form a plurality of search keyword and released product information pairs, with each search keyword and released product information pair comprising a respective search keyword and the released product information; determining and matching a plurality of features for the plurality of search keyword and released product information pairs according to a constructed first decision tree; and determining respective correlation classes of the plurality of search keyword and released product information pairs based at least in part on a result of determining and matching of the plurality of features. The disclosed method and apparatus are able to accurately and conveniently determine a matching degree between a search keyword and released product information.