Knowledge Graphs for Chatbot Query Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing chatbots, especially those based on rule-based systems, struggle to handle complex and unexpected user queries effectively, leading to limited usefulness and value for users.
Innovation Solution
A knowledge-graph system that utilizes large language models (LLMs) to generate and build knowledge graphs, allowing chatbots to answer queries by first checking the knowledge graph and prompting the LLM for novel queries, thereby improving efficiency and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If rule-based systems are used for chatbots, then the system complexity is low and ease of operation is high, but the ability to handle complex and unexpected queries is limited
Solution Approach 1:
The patent introduces a knowledge graph as an intermediary component between the user and the LLM. The knowledge graph stores structured information and relationships, allowing the chatbot to first query the knowledge graph for answers before invoking the LLM. This mediator enables the system to handle complex queries effectively while maintaining reasonable system complexity by only using the LLM when necessary.
2Adaptability or versatility
If LLMs are used to answer all queries, then the ability to handle complex queries is improved, but the time and computational resources required increase significantly
Solution Approach 1:
The system performs preliminary action by building and maintaining a knowledge graph in advance with structured information and relationships. When a user query arrives, the system first queries the pre-built knowledge graph before invoking the LLM. This preliminary preparation allows the system to answer simple queries quickly from the knowledge graph without involving the time-consuming LLM, while still being able to handle complex queries that require LLM's reasoning capabilities.
3Reliability
If LLMs are used to generate answers, then the correctness and reliability of answers improve, but bugs and inaccuracies in LLM answers still occur
Solution Approach 1:
The system implements feedback by comparing LLM-generated answers against the structured knowledge graph. When the LLM provides an answer, the system verifies it against the knowledge graph's structured information and relationships. This feedback mechanism helps identify and correct inaccuracies in LLM answers, improving overall answer reliability by leveraging the grounded, structured knowledge in the knowledge graph to validate and refine LLM outputs.
Data Source
AI summary
Techniques for a knowledge-graph system to use large language models (LLMs) to build knowledge graphs to answer queries submitted to a chatbot by users. The knowledge-graph system builds the knowledge graph using answers produced by an LLM for novel queries. The chatbot will continue to use the LLM to answer novel queries, but the chatbot may harness the knowledge graph to answer repeat questions to gain various efficiencies over LLM-backed chatbots. For example, the knowledge-graph system may easily debug or otherwise improve the answers in knowledge graphs, store provenance information in knowledge graphs, and augment the knowledge graphs using other data sources. Thus, the reliability and correctness of chatbots will be improved as the bugs and inaccuracies in answers provided by the LLM will be corrected in the knowledge graphs, but the chatbots can still harness the abilities of LLMs to provide answers across various subject-matter domains.