Automated Support Dialogue Intent Classification via Transitive Mapping

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Virtual support agents often misclassify user dialogues initially, leading to unnecessary questions and inefficiencies in resolving user issues, as they may not accurately determine user intent throughout the conversation.

Innovation Solution

A computer-implemented method that determines transitive relationships between dialogues to streamline support interactions by updating question mappings based on historical dialogues, eliminating non-relevant questions, and applying these updates to subsequent dialogues to improve intent determination and reduce interaction length.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the automated support agent asks multiple follow-up questions to determine user intent, then the accuracy of intent determination improves, but the length of the dialogue and time required increases

Engineering Contradiction:

Improveuser intent determination accuracyVSAvoiddialogue length

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary analysis by examining historical dialogues and determining transitive relationships between them before the actual support dialogue occurs. Question mappings are pre-computed based on patterns learned from previous interactions, allowing the agent to skip unnecessary questions and directly ask the most relevant ones, thus reducing dialogue length while maintaining accuracy

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system continuously learns from historical dialogues by analyzing transitive relationships and updating question mappings based on feedback from past interactions. This feedback mechanism allows the system to refine its intent determination process, eliminating questions that are proven unnecessary and improving overall efficiency over time

Inventive Principle:

Principle #23Feedback

2Reliability

If the automated support agent asks comprehensive follow-up questions to ensure accurate issue classification, then the reliability of issue resolution improves, but the number of questions and complexity of the dialogue increases

Engineering Contradiction:

Improveissue classification accuracyVSAvoiddialogue complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The system extracts only the essential and relevant questions from historical dialogues by analyzing transitive relationships. Instead of asking all possible questions, the system identifies and removes non-relevant questions based on patterns learned from previous successful resolutions, thereby reducing dialogue complexity while maintaining reliable issue classification

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

Question mappings are pre-computed and optimized based on historical dialogue analysis before actual support interactions occur. This preliminary preparation allows the system to have ready-to-use, optimized question sequences that reduce complexity while ensuring reliable classification

Inventive Principle:

Principle #10Preliminary action

3Ease of manufacture

If the automated support agent uses a fixed question mapping for all dialogues, then the ease of implementation improves, but the adaptability to different user intents and issues decreases

Engineering Contradiction:

Improvesystem implementation simplicityVSAvoiddialogue adaptation capability

Core Design Contradiction:

Ease of manufactureVSAdaptability or versatility

Solution Approach 1:

The system implements dynamic question mappings that automatically adapt to different user intents and issues. By continuously analyzing transitive relationships in historical dialogues, the system updates its question mappings to reflect learned patterns, making the dialogue process dynamic and adaptive rather than static and rigid

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system performs self-updating of question mappings by automatically analyzing historical dialogues and learning from them without requiring manual intervention. This self-service capability allows the system to improve its adaptability autonomously while maintaining implementation simplicity

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS10430466B2Streamlining support dialogues via transitive relationships between different dialogues

Publication Date: 2019.10.01 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

A computer-implemented method includes storing, by a computing device, a plurality of dialogs between user devices and an automated support application hosted by the computing device; determining, by the computing device, transitive relationships between the plurality of dialogs; and updating, by the computing device, a question mapping based on the determining the transitive relationships; and applying, the computing device, the updated question mapping to a subsequent support dialog.