Chatbot Expert Selection via Question-User Vector Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The vast amount of information available on the internet makes it difficult for users to locate relevant data, as existing technologies like search engines and digital personal assistants struggle to identify and utilize training data effectively, leading to challenges in determining trustworthy reviews and finding accurate answers to user questions.
Innovation Solution
A chatbot system that uses a knowledge base to look up known answers and, when necessary, poses questions to selected experts based on analysis of the question and user characteristics, employing a distance model to determine the best experts to answer, and trains the model using training items to improve answer selection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If search engines and digital personal assistants are used to locate information, then users can access vast amounts of data, but it becomes difficult to determine which reviews to trust and find accurate answers
Solution Approach 1:
The patent introduces an intermediary system that acts as a mediator between users and the vast amount of information available online. This system uses machine learning models to evaluate and filter information sources, providing users with reliable answers while handling the complexity of information verification behind the scenes.
Solution Approach 2:
The system implements feedback mechanisms where user interactions and outcomes are used to continuously improve the machine learning models. This feedback loop enhances the system's ability to identify trustworthy information sources and improve answer accuracy over time.
2Adaptability or versatility
If digital personal assistants use training data to improve capabilities, then they can learn and use updated information, but it challenges to identify and use training data effectively
Solution Approach 1:
The system performs preliminary actions by pre-processing and evaluating potential training data sources before they are needed. Machine learning models are pre-trained on curated datasets, and the system establishes criteria for identifying quality training data in advance, reducing the complexity of real-time training data identification.
3Measurement precision
If the chatbot system queries multiple experts to answer questions, then answer accuracy improves, but response time and system complexity increase
Solution Approach 1:
The system applies partial action by querying only a subset of potential experts rather than all possible sources. The machine learning model predicts which experts are most likely to provide accurate answers and queries only those, achieving high accuracy without the time cost of querying everyone.
Solution Approach 2:
The system changes parameters dynamically by adjusting the number of experts queried based on question complexity, time constraints, and confidence levels. This allows the system to optimize between accuracy and response time by modifying its operational parameters in real-time.
Data Source
AI summary
A chatbot can use a knowledge base including question/answer pairs to respond to questions. When a question is asked that does not correspond to a question/answer pair in the knowledge base, the chatbot can send the question to one or more humans to obtain an answer. However, only some people will have the experience, context, knowledge, etc., to answer the question. A model can be trained to select “experts” that are likely to be able to provide a good answer to a question by using both A) a vector comprising characteristics of questions and of the person posing the questions and B) a vector comprising characteristics of a possible expert. The model can trained to produce a value predicting how good an identified expert's answer is likely to be. The model can be trained based on measures of past answers provided for types of questions/questioners.