Expanding Knowledge Graphs via External Data Sources
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional question answering systems are limited by their reliance on text data within passages and knowledge graphs, which restricts the potential usefulness of the information retrieved, as they do not effectively utilize external data sources to expand entity relations and enhance candidate answer scoring.
Innovation Solution
The approach involves expanding knowledge graphs by adding new entities and relations from external data sources, such as online encyclopedias, and computing similarity scores to boost candidate answer rankings, thereby incorporating additional relevant information and improving the accuracy of question answering systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional QA pipeline is used to process passages, then the system can identify candidate answers from text data, but the knowledge graph is limited to entities and relations found only in the passage, reducing the potential usefulness of retrieved information
Solution Approach 1:
The patent merges the original knowledge graph with external knowledge graphs from multiple data sources. The system combines entity and relation information from the passage with additional entities and relations from external sources, creating an expanded knowledge graph that preserves information from both sources while eliminating duplicates through merging operations.
Solution Approach 2:
The patent introduces an intermediary process that queries external knowledge graphs using entities extracted from the passage. This intermediary mechanism retrieves additional information from external sources and integrates it into the original knowledge graph, enabling the system to leverage external knowledge without directly processing all external data.
2Quantity of substance
If knowledge graph is expanded with external data sources, then more comprehensive entity relations are obtained, but the system complexity increases
Solution Approach 1:
The patent segments the knowledge graph expansion process into distinct modules: extracting entities from passages, querying external knowledge graphs, retrieving additional entities and relations, and merging results. This segmentation allows the system to handle complex operations in manageable steps while maintaining overall system organization.
Solution Approach 2:
The patent implements a universal knowledge graph expansion mechanism that can query multiple external data sources using the same basic process. The system uses a unified approach to handle different external sources (encyclopedias, databases, etc.) through consistent entity extraction and graph querying operations, reducing the need for source-specific handling complexity.
Data Source
AI summary
An approach is provided that selects an original entity from an original knowledge graph. The approach then accesses a data source that is external to the original knowledge graph, such as an online encyclopedia. An entity in the data source is identified based on the entity matching the original entity. A new relation is then identified in the data source between the identified entity and a new entity with the new entity being absent from the original knowledge graph. An expanded knowledge graph is then generated with the expanded knowledge graph formed by adding the new entity to the original knowledge graph.