Entity Linking Using Context Keywords for Polysemous Search Terms
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional entity linking methods struggle with increased time and accuracy issues when discriminating between hundreds to thousands of named entities, particularly in search engines where polysemous words are prevalent, leading to poor performance.
Innovation Solution
A method and apparatus that extracts entities from an input context, determines candidate entities, generates keyword-based entity information, and calculates similarities to identify a top-matching entity using token embeddings and neural networks for accurate entity disambiguation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional entity linking methods using feature vectors and SVM are used, then the method is simple to implement, but the accuracy deteriorates significantly when discriminating between hundreds to thousands of named entities
Solution Approach 1:
The patent changes the parameter representation from traditional feature vectors to keyword-based entity information derived from input context. By extracting keywords from the input context and using them to generate entity information, the system achieves higher accuracy in discriminating between numerous named entities while maintaining manageable complexity through structured parameter transformation.
2Productivity
If conventional entity linking methods are used, then the implementation is straightforward, but the time required increases and accuracy decreases when handling large numbers of named entities
Solution Approach 1:
The patent performs preliminary keyword extraction from the input context before entity linking. By pre-processing the input context to identify and extract keywords, the system prepares entity information in advance, which speeds up the subsequent entity linking process while maintaining high accuracy in discriminating between large numbers of named entities.
3Measurement precision
If conventional arts are used in search engines with multiple polysemous words, then the system is simple, but entity linking accuracy becomes significantly poor
Solution Approach 1:
The patent extracts keywords from the input context that are relevant to polysemous word disambiguation. By taking out and isolating these key keywords, the system can focus entity linking operations on the most discriminative features, improving accuracy in resolving polysemous words while reducing processing time by ignoring irrelevant information.
Data Source
AI summary
Disclosed is an entity linking method. A method includes: extracting an entity from an input context including text stored in a memory; obtaining candidate entities corresponding to, and based on, the extracted entity; determining a keyword based on the input context; generating keyword-based entity information based on the keyword and based on the extracted entity; and determining a top-matching entity corresponding to the entity based on the keyword-based entity information and the plurality of candidate entities.