Question Answering System Using Source Credibility and Conversation Entropy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current knowledge graph-based systems for answering user queries face challenges in providing credible and efficient responses due to the need for additional user input and the variability of data sources, which affects the accuracy and reliability of the information provided.
Innovation Solution
The system incorporates credibility scores for data sources and minimizes user interaction by selecting the most credible path that requires the least amount of user data, using a directed acyclic graph (DAG) to determine the necessary information and optimize conversation entropy, thereby enhancing the accuracy and efficiency of query responses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the system uses multiple data sources to answer user queries, then the coverage and completeness of information is improved, but the credibility and reliability of the information decreases due to variability among data sources
Solution Approach 1:
The patent applies local quality by assigning different credibility scores to different data sources based on their specific characteristics and reliability. Each data source is evaluated individually and assigned a weight, allowing the system to treat different sources differently rather than uniformly. This enables the system to maintain high information coverage while ensuring that more credible sources have greater influence on the final answer.
Solution Approach 2:
The patent changes the parameter of data source credibility by introducing credibility scores as a quantitative parameter. This parameter is used to weight and prioritize different data sources when generating answers. By transforming the qualitative assessment of data source reliability into a measurable parameter, the system can systematically optimize for both coverage and credibility.
2Reliability
If the system requests additional user input to improve answer accuracy, then the reliability of the response is improved, but the user interaction complexity and time required increases
Solution Approach 1:
The patent applies preliminary action by pre-evaluating and storing credibility scores for multiple data sources before they are needed for answering queries. The system also pre-identifies which data sources can provide answers to potential questions. When a user query is received, the system can immediately select from pre-vetted sources without requiring additional user input, thus maintaining accuracy while minimizing interaction complexity.
Solution Approach 2:
The system performs self-service by automatically selecting the most credible data sources and generating answers without requiring user verification or additional input. The credibility scoring mechanism enables the system to autonomously determine which sources to trust, eliminating the need for users to manually verify information or provide additional constraints.
3Reliability
If the system explores multiple dialog paths to find the most credible answer, then the answer credibility is improved, but the conversation entropy and number of user prompts increases
Solution Approach 1:
The patent applies preliminary action by pre-computing credibility scores for multiple data sources and pre-identifying which sources can answer which types of questions. This allows the system to efficiently select the most credible answer path without having to explore multiple dialog paths in real-time, thus maintaining high answer credibility while minimizing conversation length and user prompts.
Solution Approach 2:
The patent replaces the mechanical process of exploring multiple dialog paths through user interaction with an automated credibility scoring and selection mechanism. Instead of manually evaluating different answer paths with user input, the system uses pre-computed credibility parameters to automatically select the optimal path, substituting automated information processing for manual exploration.
Data Source
AI summary
Systems, methods, and devices for performing interactive question answering using data source credibility and conversation entropy are disclosed. A speech-controlled device captures audio including a spoken question, and sends audio data corresponding thereto to a server(s). The server(s) performs speech processing on the audio data, and determines various stored data that can be used to determine an answer to the question. The server(s) determines which stored data to use based on the credibility of the source from which the stored data was received. The server(s) may also determine a number of user interactions needed to obtain data in order to fully answer the question and may select a question for a dialog soliciting further data based on the number of user interactions.