Feedback-Updated Chatbot Query Refinement

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Chatbots trained with insufficient or outdated data struggle to address domain-specific inquiries and generate effective structured language queries, especially in rapidly changing technological domains, and fail to adapt to user feedback in real-time.

Innovation Solution

A computer-implemented method that uses a Machine Learning model to generate primary database queries based on user inquiries, applies Natural Language Processing to adjust these queries based on user feedback, and updates the model with revised training data to improve chatbot responses dynamically, incorporating feedback from multiple user interactions and adapting to user-specific or generic feedback.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If a chatbot uses a static Machine Learning model trained on fixed training data, then the model structure is simple and easy to maintain, but the chatbot cannot adapt to domain-specific inquiries in rapidly changing technological domains

Engineering Contradiction:

Improveadaptability to domain-specific inquiriesVSAvoidmodel update mechanism complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent implements a dynamic model update mechanism where the Machine Learning model is continuously refined through feedback loops. The system transitions from a static model to a dynamic one that automatically incorporates new domain-specific information and user feedback, allowing the chatbot to adapt to rapidly changing technological domains without manual retraining

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent introduces a feedback mechanism where user responses and performance metrics are fed back into the system to update the Machine Learning model. This feedback loop enables the model to learn from actual usage patterns and improve its performance in domain-specific inquiries over time, resolving the contradiction between adaptability and complexity

Inventive Principle:

Principle #23Feedback

2Adaptability or versatility

If a chatbot uses a static Machine Learning model, then the system is simple to deploy, but the chatbot fails to adapt to user feedback in real-time

Engineering Contradiction:

Improvereal-time adaptation to user feedbackVSAvoidresponse accuracy

Core Design Contradiction:

Adaptability or versatilityVSProductivity

Solution Approach 1:

The patent implements real-time feedback mechanisms where user responses are immediately processed and used to adjust the Machine Learning model's predictions. This feedback loop enables the system to learn from user interactions in real-time, improving response accuracy dynamically without requiring system redeployment

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system performs preliminary model updates based on accumulated feedback before generating final responses. This preliminary action allows the model to pre-adjust its parameters based on user preferences and patterns, improving real-time response accuracy without adding significant system complexity

Inventive Principle:

Principle #10Preliminary action

3Reliability

If a chatbot uses insufficient or outdated training data, then the model is easier to train and deploy, but the chatbot cannot generate effective structured language queries

Engineering Contradiction:

Improvequery generation effectivenessVSAvoidmodel training time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent implements continuous model training where the Machine Learning model is continuously updated with new data from domain-specific sources and user interactions. This continuous action ensures the model maintains high reliability in generating effective queries without requiring lengthy periodic training cycles, as updates are incremental and ongoing

Inventive Principle:

Principle #20Continuity of useful action

Solution Approach 2:

The system performs preliminary data collection and model warm-up using domain-specific data before handling actual user queries. This preliminary action ensures the model has sufficient training data ready to generate effective queries immediately, reducing the time needed for full training while maintaining high reliability

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS11775511B2Feedback-updated data retrieval chatbot

Publication Date: 2023.10.03 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

A computer retrieves data from a database. The computer retrieves a Machine Learning (ML) model trained to generate database queries. The computer applies the ML model to generate a primary database query based, at least in part, on a user inquiry available to the computer. The computer retrieves the primary database query, an initial set of data from a database available to the computer. The computer, in response to retrieving the initial set of data, receives feedback assessing the initial set of data. The computer, in response to receiving the feedback, applies a Natural Language Processing (NLP) model to identify query adjustment content within the feedback. The computer revises the primary database query based, at least in part, on the model adjustment content, to generate a secondary database query. The computer retrieves using the secondary database query, a secondary set of data from the database.