Automated Fact Checking via Iterative Knowledge Graph Queries
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current fact-checking methods are time-consuming and prone to human error, struggling to keep pace with the rapid spread of information on social media platforms, and face challenges in parsing and understanding complex claims.
Innovation Solution
An automated system using iterative knowledge graph querying decomposes natural language assertions into questions and answers, translating them into structured queries to retrieve confidence scores, iteratively refining queries based on response similarity until a threshold is met, generating an assertion correctness score.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual fact checking is performed, then accuracy can be maintained, but time consumption increases and productivity decreases
Solution Approach 1:
The patent replaces manual fact-checking operations with an automated system that uses natural language processing to parse claims, translates them into knowledge graph queries, and automatically evaluates truthfulness based on retrieved knowledge. This substitution eliminates human labor while maintaining accuracy through systematic query-response evaluation.
Solution Approach 2:
The system performs self-service by automatically querying knowledge graphs, evaluating response confidence scores, and determining fact correctness without requiring human intervention. The iterative querying process enables the system to self-correct and refine results based on confidence thresholds.
2Reliability
If manual expert consultation is used for fact checking, then reliability improves, but time consumption increases
Solution Approach 1:
The patent replaces manual expert consultation with an automated knowledge graph querying system that systematically searches for factual information. The system parses claims into questions, queries knowledge graphs, and evaluates answers based on confidence scores, eliminating the need for human experts while maintaining reliable fact checking.
3Productivity
If automated fact checking systems are implemented, then productivity increases, but complexity of the system increases
Solution Approach 1:
The patent introduces a knowledge graph as an intermediary structure that bridges the gap between natural language claims and factual verification. The knowledge graph serves as a standardized representation that simplifies the complex task of fact checking by providing a structured way to store, query, and evaluate information.
Solution Approach 2:
The system segments the fact-checking process into distinct modules: claim parsing, question generation, knowledge graph querying, confidence score evaluation, and correctness determination. This segmentation reduces overall system complexity by making each component independent and easier to develop, maintain, and optimize.
4Measurement precision
If iterative querying of knowledge graphs is performed, then accuracy of fact checking improves, but processing time increases
Solution Approach 1:
The patent implements a feedback mechanism where the system evaluates confidence scores from knowledge graph queries and uses this feedback to determine whether further iterative querying is necessary. The feedback loop compares retrieved answers against the original claims and adjusts the querying process accordingly, stopping when sufficient confidence is achieved.
Solution Approach 2:
The system performs partial querying by stopping the iterative process when confidence scores meet a predetermined threshold, rather than exhaustively querying all possible knowledge graph paths. This partial action approach balances accuracy requirements with processing time constraints.
Data Source
AI summary
An embodiment includes decomposing a natural language assertion into a natural language question and answer pair that includes an initial question and an initial answer. The embodiment translates the initial question into a structured knowledge graph query and then performs an iterative process comprising iterative querying of a knowledge graph and evaluating of corresponding query responses resulting in respective confidence scores. A first iteration of the iterative process comprises querying of the knowledge graph to retrieve a first predicted answer, then determining whether a degree of similarity between the initial answer and the first predicted answer meets a threshold criterion. If not, the first predicted query is altered and used for querying the knowledge graph in a subsequent iteration of the iterative process. The embodiment also generates an assertion correctness score indicative of a degree of confidence that the assertion is factual using the respective confidence scores.