Entity Type Identification Model Contextualization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional named entity recognition (NER) systems face challenges in accurately identifying and disambiguating entity types due to their reliance on outdated dictionaries and context-free approaches, leading to incorrect or ambiguous results, especially with newly created or non-standardized entities, and are prone to overfitting with machine learning-based models.
Innovation Solution
A computer-implemented method and system that uses a machine learning-based entity type identification model to contextualize entity names by considering the surrounding text, minimizing overfitting and improving accuracy by coupling with NER systems to provide more robust entity type prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional NER systems use manually curated entity dictionaries, then they can identify known entities, but they cannot adapt to new or non-standardized entities and require continual manual updates
Solution Approach 1:
The system enables automatic entity type identification through machine learning models that self-adapt to new entities without manual intervention. The model learns from training data and automatically updates its knowledge base, eliminating the need for manual dictionary curation while maintaining high accuracy in identifying both known and new entities.
2Measurement precision
If NER systems use context-free text matching, then they can process text quickly, but they produce incorrect results when entities have the same name but different types
Solution Approach 1:
The system transitions from context-free text matching to context-aware entity type identification by incorporating surrounding text as an additional dimension. The machine learning model analyzes both the entity name and its contextual environment, enabling accurate disambiguation of entities with identical names but different types through contextual features.
3Measurement precision
If machine learning models are used for entity type identification, then they can handle context, but they become prone to overfitting
Solution Approach 1:
The system applies partial contextual information rather than using excessive or complete context. By selectively incorporating relevant surrounding text features and limiting the scope of contextual analysis, the model achieves good contextual understanding while avoiding overfitting to specific training examples, thus maintaining generalization capability.
Data Source
AI summary
Method(s), apparatus and system(s) are provided for entity type identification and/or disambiguation of entities within a corpus of text the method including: receiving one or more entity results, each entity result comprising data representative of an identified entity and a location of the identified entity within the corpus of text; identifying an entity type for each entity of the received entity results by inputting text associated with the location of said each entity in the corpus of text to a trained entity type (ET) model configured for predicting or extracting an entity type of said each entity from the corpus of text; and outputting data representative of the identified entity type of each entity in the received entity results.