Conversation Efficiency Score via Base Expansion Pair Ratio

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current natural language conversation systems face challenges in accurately measuring mutual understanding between users and machines, as they rely on complex dialogue acts and lack efficient methods for analyzing conversational actions and repairs, leading to suboptimal interaction efficiency.

Innovation Solution

A computer-implemented method and system that analyzes natural language conversations by determining base and expansion pairs, computing an efficiency score as a ratio of base pairs to expansion pairs, and using a labeling scheme to automatically label utterances based on adjacency pair positions, facilitating improved classifier performance and mutual understanding.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If current dialogue act labeling methods are used to analyze conversational actions, then the system can detect user intent, but the complexity of the analysis process increases and measurement precision of mutual understanding deteriorates

Engineering Contradiction:

Improvemeasurement precision of mutual understandingVSAvoidcomplexity of dialogue act labeling

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the conversation analysis into distinct adjacency pair types (base pairs, expansion pairs, repair pairs) rather than using complex dialogue act labels. This segmentation simplifies the classification process while maintaining measurement precision by focusing on structural conversation patterns that directly indicate mutual understanding.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent extracts and focuses on specific structural elements of conversation (adjacency pairs and their relationships) while eliminating the need for complex dialogue act labeling. By taking out only the essential structural components needed to measure mutual understanding, the system achieves simpler analysis with improved measurement precision.

Inventive Principle:

Principle #2Taking out (Extraction)

2Productivity

If detailed dialogue act labeling is performed on all utterances, then comprehensive conversation analysis is achieved, but the time and computational resources required increase

Engineering Contradiction:

Improveconversation analysis efficiencyVSAvoidtime for analyzing conversational actions

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The patent segments utterances into adjacency pairs and categorizes them by type (base, expansion, repair) rather than applying detailed dialogue act labels to every utterance. This segmentation enables efficient processing by focusing on the structural relationships that matter most for measuring mutual understanding, significantly reducing analysis time.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies partial action by focusing only on the essential structural analysis of adjacency pairs rather than performing complete dialogue act labeling on all utterances. This partial approach achieves sufficient measurement precision for mutual understanding while dramatically improving productivity and reducing time loss.

Inventive Principle:

Principle #16Partial or excessive action

3Measurement precision

If the system uses traditional conversation analysis methods, then it can process natural language input, but the ability to accurately measure mutual understanding deteriorates

Engineering Contradiction:

Improveaccuracy of mutual understanding measurementVSAvoidease of analyzing conversational flow

Core Design Contradiction:

Measurement precisionVSEase of operation

Solution Approach 1:

The patent introduces adjacency pair analysis as an intermediary framework between traditional conversation processing and mutual understanding measurement. This intermediary structure provides a systematic way to analyze conversational flow while accurately measuring mutual understanding through the relationships between base pairs, expansion pairs, and repair pairs.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent shifts the analysis from traditional dialogue act dimensions to a structural dimension based on adjacency pair relationships. By analyzing conversations in this new dimensional framework, the system achieves both ease of operation through systematic categorization and high measurement precision for mutual understanding.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

Data Source

PatentUS10303772B2Measuring mutual understanding in human-computer conversation

Publication Date: 2019.05.28 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

Technical solutions are described method for analyzing a natural language conversation-generating machine. An example computer implemented method includes determining, from a plurality of adjacency pairs in a conversation, a number of base pairs. The computer implemented method also includes determining, from the plurality of adjacency pairs, a number of expansion pairs. The computer implemented method also includes computing a efficiency score for the conversation by computing a ratio of the number of base pairs and the number of expansion pairs.