Causal Hypothesis Scoring with Sentiment Propagation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Machine-computed scores for hypotheses about individual events are less reliable due to limited relevant data, making it challenging to accurately rank argument statements and predict outcomes based on causal relationships.
Innovation Solution
A computer-implemented method that creates a causal relationship model, propagates pro and con sentiment scores from leaf hypotheses to a root hypothesis using axioms, and merges general and specific sentiment scores to provide a final prediction, ensuring reasonableness and reliability of the scoring.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If machine-computed scores are generated for hypotheses about individual events, then specific predictions can be made, but the reliability of the scores decreases due to limited relevant data
Solution Approach 1:
The patent introduces a causal relationship model as an intermediary structure that connects individual event hypotheses to general event hypotheses. This mediator allows information to flow from general hypotheses (with abundant data) to specific hypotheses (with limited data), improving the reliability of individual event predictions without sacrificing the ability to make specific predictions
Solution Approach 2:
The patent adds a new dimension to the scoring system by creating a hierarchical structure with multiple levels (individual events, general events, and causal relationships). This dimensional expansion allows the system to leverage data from multiple levels, transforming the reliability problem from a single-dimension constraint into a multi-dimension solution space
2Ease of manufacture
If only attack relationships are used for scoring claim statements, then the scoring process is simple, but the accuracy of ranking argument statements decreases
Solution Approach 1:
The patent merges multiple types of relationships (attack, support, and causal relationships) into a unified scoring framework. By combining these different relationship types, the system achieves more accurate ranking of argument statements while maintaining a relatively simple scoring process through the use of a structured model
3Measurement precision
If sentiment analysis is performed on individual event hypotheses, then specific predictions are obtained, but the quality of scores decreases due to less relevant data
Solution Approach 1:
The patent performs preliminary sentiment analysis on general event hypotheses first, before analyzing individual event hypotheses. This preliminary action establishes a foundation of reliable scores from general events that can then be used to inform and improve the scoring of specific individual events, even when data for those individual events is limited
Data Source
AI summary
A computer implemented method of hypothesis scoring based on causal relationships is provided. The computer implemented method includes creating a causal relationship model utilizing a plurality of hypotheses and a causal relationship between each of two or more pairs of hypotheses, and obtaining pro and con sentiment scores for each hypothesis utilizing a scoring function. The computer implemented method further includes assigning the obtained pro and con sentiment scores to each hypothesis in the causal relationship model, and propagating the pro and con sentiment scores from leaf hypotheses to a root hypothesis utilizing axioms to test the propagating scores for reasonableness. The computer implemented method further includes determining a final pro and con score for the root hypothesis, and presenting the final pro and con scores representing a prediction of the hypotheses to a user.