Chat Transcript Triplet Analysis for Response Matching

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Customer representatives in live chat interfaces often lack the necessary skills to address diverse customer questions efficiently, leading to increased response time and potential dissatisfaction, as they may need to transfer customers to more skilled representatives based on varying business priorities.

Innovation Solution

A method that analyzes historical chat transcript data to generate multi-dimensional success vectors, tagging intents, entities, and sentiments, which are used to rate customer representative responses and match customers with the most qualified representatives based on past performance, improving response effectiveness and customer satisfaction.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If customer representatives are educated on only a portion of questions, then training time and resources are reduced, but response effectiveness and customer satisfaction deteriorate

Engineering Contradiction:

Improveresponse effectivenessVSAvoidtraining time

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The system pre-analyzes historical chat transcripts to extract questions, answers, and success factors before they are needed. By preparing and storing this knowledge in advance, the system enables rapid retrieval during live chats without requiring extensive real-time analysis or prolonged training sessions for representatives.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system creates simplified copies of successful customer representative responses from historical data. These copied responses, along with their associated success vectors and metadata, are stored and can be quickly referenced during live chats, allowing representatives to leverage proven solutions without memorizing extensive content.

Inventive Principle:

Principle #26Copying

2Measurement precision

If customer representatives search for appropriate answers, then answer accuracy may improve, but response time increases

Engineering Contradiction:

Improveanswer accuracyVSAvoidresponse time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system incorporates success vectors derived from historical chat outcomes as feedback signals. By analyzing whether past responses led to successful resolutions, the system refines its recommendations and provides accuracy indicators, allowing representatives to quickly identify high-quality answers without extensive searching.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system pre-processes historical data to identify and tag successful response patterns, intents, and entities before they are needed. This preliminary analysis creates a ready-to-use knowledge base that enables rapid accurate response retrieval during live chats.

Inventive Principle:

Principle #10Preliminary action

3Reliability

If customers are transferred to second customer representatives, then issue resolution effectiveness improves, but customer waiting time and operational complexity increase

Engineering Contradiction:

Improveissue resolution effectivenessVSAvoidcustomer waiting time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system pre-analyzes customer questions and matches them with suitable representatives based on historical success data before the chat begins. By preparing match recommendations in advance using multi-dimensional success vectors, the system enables immediate routing to the most qualified representative, eliminating transfer delays.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system replaces the manual process of customer transfers with an automated matching mechanism. Using machine learning models and success vectors, the system automatically identifies and routes customers to appropriate representatives based on question intent, entities, and representative expertise, eliminating the need for manual transfers and reducing waiting time.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

4Measurement precision

If chat transcripts are analyzed in detail, then response quality improves, but processing complexity and computational resources increase

Engineering Contradiction:

Improveresponse quality measurementVSAvoidprocessing complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system segments chat transcripts into discrete triplets (question, answer, outcome) and further divides them into tagged components such as intents, entities, and success factors. This segmentation allows for efficient processing and analysis of specific elements without requiring complex analysis of entire transcripts, reducing computational complexity while maintaining measurement precision.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS11210677B2Measuring the effectiveness of individual customer representative responses in historical chat transcripts

Publication Date: 2021.12.28 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

A method, computer system, and a computer program product for response effectiveness is provided. The present invention may include receiving a chat transcript. The present invention may include separating the chat transcript into a set of triplets, the set including two or more triplets. The present invention may include tagging each triplet in the set of triplets with one or more tags, wherein the one or more tags includes an intent, an entity, and a sentiment. The present invention may include generating at least one multi-dimensional success vector. The present invention may include aggregating the generated multi-dimensional success vectors to determine an overall satisfaction.