Dynamic Entity Extraction for Knowledge Base Currency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional knowledge bases often fail to provide up-to-date or reliable information regarding entities, as they struggle to dynamically extract and store information from electronic documents, leading to outdated search results and reduced search query efficiency.
Innovation Solution
A computer-implemented system and method that detects new entities in electronic documents, determines their classes and attributes based on surrounding lexical contexts and known entities, and dynamically updates these associations in real-time or near real-time, using a database to store and retrieve this information for improved search engine functionality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional knowledge bases are used to store entity information, then data structure stability is maintained, but information up-to-date reliability deteriorates
Solution Approach 1:
The patent implements dynamic entity extraction and classification systems that continuously update knowledge bases in real-time as new documents are processed. The system transitions from static, periodic updates to dynamic, event-driven updates triggered by new document ingestion, ensuring knowledge bases reflect current information while maintaining structural integrity through controlled update mechanisms.
Solution Approach 2:
The system employs feedback loops where extracted entity information is continuously validated, classified, and integrated back into the knowledge base. The classification models are refined based on feedback from newly processed documents, creating a self-improving system that maintains reliability while adapting to new information patterns.
2Measurement precision
If manual entity classification is performed, then classification accuracy is improved, but processing speed deteriorates
Solution Approach 1:
The patent introduces machine learning classification models as intermediaries between raw entity extraction and final knowledge base storage. These models automatically classify entities into predefined categories (e.g., persons, organizations, locations, events) with high accuracy, eliminating the need for manual classification while maintaining or improving precision compared to human reviewers.
Solution Approach 2:
The system replaces manual, mechanical classification processes with automated computational classification models. The models use natural language processing and machine learning algorithms to perform classification tasks that were previously done manually, achieving both high accuracy and scalable processing speeds.
3Reliability
If frequent knowledge base updates are performed, then information currency is improved, but system complexity deteriorates
Solution Approach 1:
The system implements periodic batch processing of documents combined with event-triggered updates. Instead of continuous real-time updates that would overwhelm the system, documents are processed in periodic batches, with knowledge base updates triggered by specific events (e.g., completion of a batch, detection of new entity types), balancing information currency with system manageability.
Solution Approach 2:
The knowledge base update process is segmented into independent, modular components: document ingestion, entity extraction, classification, validation, and integration. Each segment can be processed independently and in parallel, reducing overall system complexity while enabling frequent updates through distributed processing.
4Loss of information
If comprehensive entity extraction is performed, then information completeness is improved, but processing time deteriorates
Solution Approach 1:
The system applies partial extraction in the first pass, focusing on high-priority entity types and attributes most relevant to search queries. Less critical entities are extracted in subsequent passes or only when triggered by specific conditions, reducing initial processing time while maintaining information completeness for essential data.
Solution Approach 2:
The extraction process applies different levels of detail to different entity types based on their importance and usage patterns. High-value entities (e.g., persons, organizations) receive comprehensive extraction with multiple attributes, while less critical entities receive minimal extraction, optimizing the balance between completeness and processing efficiency.
Data Source
AI summary
Computer-implemented systems and methods are provided for extracting and storing information regarding entities from documents, such as webpages. In one implementation, a system is provided that detects an entity candidate in a document and determines that the detected candidate is a new entity. The system also detects a known entity proximate to the known entity based on the one or more entity models. The system also detects a context proximate to the new and known entities having a lexical relationship to the known entity. The system also determines a second entity class associated with the known entity and a context class associated with the context. The system also generates a first entity class based on the second entity class and the context class. The system also generates an entry in the one or more entity models reflecting an association between the new entity and the first entity class.