Automated Call Segmentation Using Speech Recognition and NLP
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current contact center call analysis techniques are limited in automatically segmenting and classifying call segments, leading to inefficiencies in data utilization and requiring expensive, time-consuming manual analysis, which restricts the ability to fully leverage call data for insights and performance improvement.
Innovation Solution
The implementation of natural language processing and machine learning technologies to automatically detect utterance boundaries, classify utterances, and partition call transcripts into segments, utilizing a system with components for utterance boundary detection, classification, and normalization, enabling the identification of predefined and industry-specific call sections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual analysis by consultants is used to identify call segments, then measurement precision of call segments is improved, but productivity and loss of time worsen due to expensive and slow manual processes
Solution Approach 1:
The patent replaces the manual mechanical process of consultants listening to and analyzing calls with an automated system using speech recognition, natural language processing, and machine learning algorithms. The system automatically transcribes calls, identifies utterance boundaries, classifies call segments, and generates statistics without human intervention, thereby maintaining measurement precision while dramatically improving productivity.
2Measurement precision
If manual analysis is used to study call segments, then measurement precision is improved, but loss of time increases making it impossible to study large volumes of calls
Solution Approach 1:
The system substitutes manual time-consuming analysis with automated processing using speech recognition and natural language processing algorithms that can analyze calls in real-time or near-real-time, reducing analysis duration from hours to minutes while maintaining classification accuracy through trained machine learning models.
Solution Approach 2:
The system performs preliminary actions by pre-training machine learning models on annotated call data and pre-processing call transcripts through speech recognition and normalization before actual analysis, enabling rapid and accurate segmentation without manual intervention during the analysis phase.
3Productivity
If automatic speech recognition and natural language processing are applied to detect utterance boundaries and classify segments, then productivity and loss of time are improved, but device complexity increases
Solution Approach 1:
The patent segments the complex automated analysis system into distinct functional modules: speech recognition module, utterance boundary detection module, call segment classification module, and statistics generation module. Each module performs a specific function and can be independently developed, tested, and maintained, reducing overall system complexity while maintaining high productivity.
Solution Approach 2:
The system introduces intermediary components such as transcript normalization layers and feature extraction modules that bridge the gap between raw speech data and final analysis results, simplifying the overall processing pipeline by breaking down complex transformations into manageable intermediate steps.
Data Source
AI summary
A system and method for automatic call segmentation including steps and means for automatically detecting boundaries between utterances in the call transcripts; automatically classifying utterances into target call sections; automatically partitioning the call transcript into call segments; and outputting a segmented call transcript. A training method and apparatus for training the system to perform automatic call segmentation includes steps and means for providing at least one training transcript with annotated call sections; normalizing the at least one training transcript; and performing statistical analysis on the at least one training transcript.