Dynamic Attribute Extraction in AI Text Mining
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing text mining engines lack the ability to understand syntactical relationships in unstructured text, preventing them from creating and reporting entity relationships.
Innovation Solution
A rules layer is added to the AI platform, enabling dynamic attribute extraction by defining relationship patterns, entities, and actions, which allows for the recognition and attachment of dynamic attributes to entities, thereby creating relationships among them.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a text mining engine extracts entities from unstructured text, then entity recognition capability is improved, but the ability to understand syntactical relationships and create entity relationships deteriorates
Solution Approach 1:
The patent segments the text mining process into distinct functional layers: a core entity recognition layer that identifies entities, and a separate rules layer that handles syntactical relationship analysis. This segmentation allows each layer to specialize in its function, with the entity recognition layer focusing on extracting entities and the rules layer focusing on understanding relationships between them, thereby resolving the contradiction between entity recognition capability and syntactical relationship understanding.
Solution Approach 2:
The patent adds a new dimensional layer (the rules layer) above the existing entity recognition layer. This additional dimension enables the system to process not only entity extraction but also syntactical relationships, sentence structure, and grammatical composition. By operating in this elevated dimension, the system can create entity relationships while maintaining the entity recognition capabilities of the underlying layer.
2Productivity
If existing text mining engines extract entities from text, then information extraction is improved, but the ability to analyze sentence structure and grammar deteriorates
Solution Approach 1:
The rules layer serves as an intermediary between the entity extraction process and the syntactical analysis process. It receives entities from the core extraction layer and mediates by applying grammatical and syntactical rules to determine relationships between entities. This intermediary layer enables the system to maintain high information extraction efficiency while simultaneously achieving sentence structure and grammar understanding through the mediation of the rules layer.
3Adaptability or versatility
If a rules layer is added to enable dynamic attribute extraction, then entity relationship creation capability is improved, but system complexity increases
Solution Approach 1:
The rules layer is designed with universal applicability, handling multiple functions including syntactical relationship analysis, dynamic attribute extraction, and entity relationship creation through a single unified layer. This multi-functionality reduces the need for separate specialized components, thereby managing system complexity while enhancing entity relationship creation capability. The rules layer can be configured with different rule sets for various domains without requiring fundamental architectural changes.
Solution Approach 2:
The rules layer incorporates dynamic characteristics, allowing the system to adapt to different text types, domains, and relationship patterns through configurable rules. This dynamic nature enables the system to handle diverse entity relationship creation tasks without requiring separate static processing modules for each scenario, thereby managing complexity through flexibility rather than through multiple fixed components.
Data Source
AI summary
An AI platform may receive a request for information on text. The text is processed through a text mining pipeline for dynamic attribute extraction. An engine determines entities in the text and utilizes the entities to determine a relationship pattern. The engine identifies a trigger by matching one of the entities with a predefined entity in a utility authority file, locates an entity in close proximity to the trigger, identifies a value or regular expression in close proximity to the trigger in the text, and creates a triplet containing the entity, the trigger, and the value or regular expression, the triplet representing the relationship pattern. The engine applies an action to the triplet, wherein the action comprises obtaining the value from the text or translating the regular expression. The engine attaches the value or a result from the translating to the entity as a dynamic attribute of the entity.