METHOD AND APPARATUS FOR DETECTION OF SENTIMENT IN AUTOMATED TRANSCRIPTIONS
First Claim
1. A method for automatically detecting sentiments in an audio signal of an interaction held in a call center, comprising:
- receiving the audio signal from a logging and capturing unit associated with the call center;
performing audio analysis on the audio signal to obtain text spoken within the interaction;
segmenting the text into context units;
extracting a sentiment candidate context unit from the context units;
extracting features from the sentiment candidate context unit or from the audio signal;
determining in accordance with the features whether the sentiment candidate context unit is valid or erroneous; and
determining sentiment polarity and intensity for the sentiment candidate context unit.
3 Assignments
0 Petitions
Accused Products
Abstract
A method and apparatus for automatically detecting sentiment in interactions. The method and apparatus include training, in which a model is generated upon features extracted from training interactions and tagging information. and run-time in which the model is used towards detecting sentiment in further interactions.
The method and apparatus train and use models for classifying context units as containing sentiment or not, the models generated upon tagging data for indicating whether the context units indeed contain sentiment, and the sentiment polarity and intensity within the context unit. A further, upper level model is generated for providing a total score for the interaction, based on the sentiment detected and the classification scores of context units within the interaction.
125 Citations
20 Claims
-
1. A method for automatically detecting sentiments in an audio signal of an interaction held in a call center, comprising:
-
receiving the audio signal from a logging and capturing unit associated with the call center; performing audio analysis on the audio signal to obtain text spoken within the interaction; segmenting the text into context units; extracting a sentiment candidate context unit from the context units; extracting features from the sentiment candidate context unit or from the audio signal; determining in accordance with the features whether the sentiment candidate context unit is valid or erroneous; and determining sentiment polarity and intensity for the sentiment candidate context unit. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14)
-
-
15. An apparatus for automatically detecting sentiments in an audio signal of an interaction held in a call center, comprising;
-
an interaction receiving component arranged to receive an audio signal representing. the interaction; a context unit segmentation component arranged to segment the audio signal into context units; a sentiment candidate identification component for identifying sentiment candidate from the context units; a feature extraction component arranged to extract at least text from a part of the audio signal associated with the sentiment candidates; and a classification component arranged to apply a model to the feature vector to determine whether the audio signal contains sentiment, or sentiment polarity and intensity within the audio signal. - View Dependent Claims (16, 17, 18, 19)
-
-
20. A computer readable storage medium containing a set of instructions for a general purpose computer, the set of instructions comprising:
-
receiving an audio signal of an interaction held in a call center from a logging and capturing unit associated with the call center; performing audio analysis on the audio signal to obtain text spoken within the interaction; segmenting the text into context units; extracting a sentiment candidate context unit from the context units; extracting features from the sentiment candidate context unit and from the audio signal; determining in accordance with the features whether the sentiment candidate context unit is valid or erroneous; and determining sentiment polarity and intensity for the sentiment candidate context unit.
-
Specification