Sentiment Detection in Automated Transcriptions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems face challenges in detecting sentiment in vocal interactions with high accuracy, as they require manual listening to a large number of calls to identify sentiment, making the process labor-intensive and inefficient.
Innovation Solution
A method and apparatus for automatically detecting sentiment in audio signals by classifying feature vectors extracted from the signals using predetermined models, including speech-to-text engines, emotion detection engines, and validity models to identify and quantify sentiment polarity and intensity in interactions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual listening to calls is used to detect sentiment, then detection accuracy can be maintained, but the process becomes labor-intensive and inefficient
Solution Approach 1:
The patent replaces the mechanical manual listening process with an automated computer-based system that performs audio analysis, text extraction, and sentiment classification using machine learning models, thereby eliminating manual labor while maintaining detection capability
Solution Approach 2:
The system enables self-service sentiment detection by automatically processing audio calls through multiple engines (audio analysis, text extraction, sentiment classification) without requiring human intervention, allowing the system to serve itself in detecting sentiments
2Productivity
If automated sentiment detection is implemented, then productivity increases, but system complexity increases
Solution Approach 1:
The patent segments the sentiment detection system into distinct functional modules: audio analysis engine, text extraction engine, sentiment classification engine with separate training and validation phases, allowing each component to be developed and optimized independently while maintaining overall system productivity
Solution Approach 2:
The system employs universal machine learning models that can handle multiple tasks including audio transcription, text extraction, and sentiment classification across different contexts, reducing the need for separate specialized systems and thereby managing complexity
3Measurement precision
If multiple analysis engines are used to improve accuracy, then measurement precision increases, but device complexity increases
Solution Approach 1:
The patent merges multiple analysis functions into an integrated sentiment detection system where audio analysis, text extraction, and sentiment classification work together as a unified pipeline, combining their capabilities to achieve high accuracy while managing complexity through integration rather than separate standalone systems
Data Source
AI summary
A method for automatically detecting sentiments in an audio signal of an interaction held in a call center, including, receiving the audio signal from a logging and capturing unit. Performing audio analysis on the audio signal to obtain text spoken within the interaction. Segmenting the text into context units according to acoustic information acquired from the audio signal to identify units of speech bound by non-speech segments, wherein each context unit includes one or more words. Extracting a sentiment candidate context unit from the context units using a phonetic based search. Extracting linguistic features from the text of the sentiment candidate context unit and acoustic features from a segment of the audio signal associated with the sentiment candidate context unit. Determining in accordance with the linguistic features and acoustic features whether the sentiment candidate context unit is valid or erroneous, and determining sentiment polarity and intensity.