Topic-Routed Dialogue System for Multi-Topic Query Accuracy

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current dialogue robots face challenges in generating accurate replies due to their reliance on large databases and limited topic coverage, often resulting in irrelevant responses, especially when user queries span multiple topics.

Innovation Solution

A human-computer dialogue method and apparatus that pre-configure multiple dialogue robots for specific topics, allowing for dynamic topic determination and allocation of user input to the most relevant robot based on predefined mapping relationships, enabling semantic understanding and accurate response generation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If a single dialogue robot relies on a large dialogue database and intelligent matching technology, then it can handle various user queries, but it produces irrelevant replies when the dialogue database is insufficient or user problems cannot be matched

Engineering Contradiction:
Improvetopic coverageVSAvoidreply accuracy
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The patent divides a single dialogue robot into multiple specialized dialogue robots, each trained on a specific topic. This segmentation allows each robot to excel in its domain while the system as a whole maintains broad coverage through topic routing based on user input.

Inventive Principle:
Principle #1Segmentation

2Ease of operation

If a generative dialogue robot uses deep learning to understand user questions and generate replies word by word, then it can produce human-like responses, but it cannot adaptively generate accurate replies when user questions span multiple topics

Engineering Contradiction:
Improvenatural language understandingVSAvoidmulti-topic handling
Core Design Contradiction:
Ease of operationVSAdaptability or versatility

Solution Approach 1:

The patent segments the generative dialogue capability across multiple specialized robots, each proficient in natural language understanding for its specific topic. The topic identification module routes queries to the appropriate specialist, enabling accurate handling of multi-topic questions.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces a topic identification module as an intermediary between the user input and dialogue robots. This mediator analyzes the user question, identifies relevant topics, and routes the query to the appropriate specialized robot, enabling accurate multi-topic handling.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Reliability

If multiple dialogue robots are configured for different topics, then accurate replies can be generated for specific topics, but the system complexity increases

Engineering Contradiction:
Improvereply accuracyVSAvoidsystem structure
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent creates a universal framework where multiple specialized dialogue robots operate under a common topic identification and routing system. This multi-functional architecture allows the system to handle various topics accurately while maintaining a unified, manageable structure through standardized interfaces and protocols.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentUS12073334B2Human-computer dialogue method and apparatus
Publication Date: 2024.08.27 HUAWEI TECH CO LTD
  • US12073334B2 patent drawing
  • US12073334B2 patent drawing
  • US12073334B2 patent drawing

AI summary

A method includes: obtaining a text entered by a user; determining at least one topic related to the text; determining a target dialogue robot from the plurality of dialogue robots based on the at least one topic related to the text and a predefined mapping relationship between a dialogue robot and a topic, where a target topic corresponding to the target dialogue robot is some or all of the at least one topic related to the text; allocating the text to the target dialogue robot; and obtaining a reply for the text from the target dialogue robot, where the reply is generated by the target dialogue robot based on at least one semantic understanding of the text.