Contact Center Taxonomy Analytics from LLM Conversation Insights
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Contact centers face challenges in efficiently analyzing large volumes of unstructured textual data from customer-agent interactions, which are noisy and time-consuming, making it difficult to derive operational insights for performance improvement.
Innovation Solution
A method using large language models (LLM) to generate a hierarchical taxonomy by processing conversation data, including generating insights, category names, and assignments, and grouping them to create a visual representation for user interface display.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional methods are used to analyze unstructured textual data from customer-agent interactions, then comprehensive analysis can be performed, but the process is time-consuming and inefficient
Solution Approach 1:
The patent replaces traditional mechanical text analysis methods with a large language model (LLM)-based automated system. The LLM processes unstructured conversation data, generates insights, creates hierarchical taxonomies, and identifies operational issues without manual intervention, dramatically improving analysis efficiency while reducing time consumption.
2Loss of information
If manual analysis of conversation data is performed, then detailed insights can be obtained, but the process becomes costly and slow
Solution Approach 1:
The system enables self-service analysis where the LLM automatically processes conversation data, generates insights, creates taxonomies, and identifies operational issues without requiring manual analysis. The automated system serves itself to extract and organize information from unstructured data, maintaining comprehensive insight generation while eliminating time-consuming manual processes.
3Reliability
If comprehensive taxonomy analysis is performed on all interaction data, then complete operational visibility is achieved, but the complexity and resource requirements increase
Solution Approach 1:
The patent segments the analysis process into distinct hierarchical levels: conversation data is first processed to generate individual insights, then insights are organized into taxonomic categories, and finally structured into a hierarchical taxonomy. This segmentation allows comprehensive analysis to be performed in manageable stages, reducing system complexity while maintaining complete operational visibility and reliable supervisor awareness.
Data Source
AI summary
A method for generating a hierarchical taxonomy relating to an interaction aspect from conversation data. The method includes a first process for generating insights that includes: receiving conversation data for a first interaction; determining a conversation portion relevant to the interaction aspect and a question prompt and providing them as inputs to an LLM; and generating responsive output text via the LLM as the first insight. In a second process, inputs are provided to the LLM that include the insights, a first instruction to generate category names based on the insights, and a second instruction to make a category assignment for each insight. The second process further includes receiving from the LLM the generated category names and category assignments; grouping the insights by those having the same category assignment; and generating a hierarchical taxonomy according to the groupings.