Information Discovery System Using NLP Knowledge Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing information retrieval systems fail to effectively discover relevant information related to a topic from various sources, as they rely on direct keyword searching, which may miss items and sources that do not include the searched keywords, making it difficult to find relevant data.
Innovation Solution
An information discovery system that uses Natural Language Processing (NLP) to extract knowledge points from data elements, such as author and recipient information, sentiment, and purchase intent, and creates a knowledge graph with weighted links between these points and data elements, allowing for traversal based on user-defined queries to uncover semantically relevant data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If direct keyword searching is used, then search speed and simplicity are improved, but information completeness and relevance are worsened
Solution Approach 1:
The patent introduces an intermediary system between the user query and the information sources. This system uses Natural Language Processing to extract knowledge points from queries and creates a knowledge graph that mediates the search process, enabling discovery of relevant information beyond direct keyword matches while maintaining search efficiency
Solution Approach 2:
The patent replaces the mechanical keyword-matching system with an intelligent NLP-based system. Instead of relying on simple string matching, the system uses semantic understanding, knowledge extraction, and graph-based relationships to identify relevant information, substituting a rigid mechanical approach with a flexible intelligent one
2Adaptability or versatility
If knowledge graph traversal is implemented, then information discovery capability is improved, but system complexity is worsened
Solution Approach 1:
The patent segments the complex information retrieval task into distinct components: NLP processing to extract knowledge points, knowledge graph construction to organize relationships, and graph traversal to discover information. This segmentation allows each component to be optimized independently while working together to solve the overall problem
Solution Approach 2:
The knowledge graph serves as an intermediary data structure that simplifies the complexity. Instead of directly querying unstructured data sources, the system uses the structured knowledge graph as a mediator, pre-computing relationships and enabling efficient traversal while hiding the underlying complexity from users
Data Source
AI summary
Systems, device and techniques are disclosed for an information discovery system. An element of data may be retrieved. A knowledge point may be extracted from the element of data. The knowledge point may include an aspect of the element of data. The element of data and the knowledge point may be linked with a traversable link. The knowledge point may further be linked to a second element of data. Natural language processing analysis, linguistic analysis, sentiment analysis, and metadata analysis, may be used to determine the aspect of the element of data.