Structured Chat Summarization via Template Ranking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Unstructured customer service chats on chat platforms contain unwanted information, making it difficult to extract meaningful insights and analyze customer service data effectively.
Innovation Solution
A method that groups digital chat records by tasks, generates task keywords and related words, and creates expandable template data structures to extract and rank chat utterances and snippets, forming structured summary data structures in human-readable and SQL-compatible formats for efficient searching.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If customer service chats are stored as unstructured human generated textual data, then the data can be easily collected and stored, but the data contains unwanted information and is difficult to analyze effectively
Solution Approach 1:
The patent segments unstructured chat data into structured components by extracting specific entities (customer names, product names, issues, resolutions) and organizing them into predefined templates. This segmentation transforms raw textual data into analyzable structured formats while maintaining ease of collection.
Solution Approach 2:
The patent introduces an intermediary processing system that includes template matching modules, entity extraction algorithms, and data normalization layers. This intermediary transforms unstructured chat data into structured formats, enabling efficient analysis without changing the original data collection method.
2Reliability
If chat data is kept in original unstructured format, then data integrity is maintained, but meaningful insights cannot be extracted effectively
Solution Approach 1:
The patent applies preliminary structuring actions to chat data by pre-defining templates for common customer service scenarios and pre-extracting key entities before full analysis. This preliminary organization preserves data integrity while making insights extractable through subsequent processing.
Solution Approach 2:
The patent changes the organizational parameters of chat data from free-form text to structured fields with specific parameters (customer_name, product_name, issue_type, resolution). This parameter transformation maintains the original information while enabling systematic analysis and insight extraction.
3Loss of information
If all chat data is analyzed in detail, then comprehensive insights are obtained, but the processing time and computational resources increase significantly
Solution Approach 1:
The patent applies partial action by focusing analysis on extracted key entities and template-matched sections rather than processing entire chat transcripts. This selective approach obtains comprehensive insights about customer service metrics while reducing processing time by ignoring redundant conversational filler.
Solution Approach 2:
The patent extracts only the essential information elements (issues, resolutions, product names) from chat data using entity extraction and template matching. This extraction approach obtains comprehensive customer service insights while minimizing processing time by excluding unnecessary textual content.
Data Source
AI summary
At least some embodiments are directed to a system to compute uniform structured summarization of customer chats. In at least some embodiments, the system may operate a processor and receive a corpus of chats between customers and customer service representatives of an enterprise. Grouping the corpus of chats into subgroup task types and then extracting chat keywords and chat related words for each subgroup task type. Generating an expandable template data structure for each subgroup task type. Processing at least one chat to extract chat utterances and chat snippets ranking the chat utterances and chat snippets. Populating the expandable template data structure based on rankings to generate a chat summary data structure.