- This technology automatically predicts important utterances to identify focus points from speech recognition results of the customer and operator’s telephone conversation, and improves accuracy based on operator input of correct information.
- This technology will reduce conversation time, improve customer satisfaction, and help operators to make contact call reports by quickly understanding the focus points of dialogue.
Background and existing issues
A major issue of a contact center is to reduce costs of work during and after a call by reducing work time with the support of computer.
Advantages of this technology
- Dialog structure prediction: Dialogues with contact centers tend to be centered on business tasks such as inquiries regarding products or requests for procedures related to a particular service, so typical patterns appear in the flow of dialogue. This technology is able to predict dialogue segments accurately DNN model trained by using call center conversation logs. Having an overall grasp of conversion flow in this way provides strong clues for understanding the dialogue between operator and customer.
- Gradually learning and improving: When the extracted main points are incorrect, the operator can correct them. Such corrected data are then used to update DNN model, and autonomously improves the prediction performance.
- Dialogue browser: Operator can efficiently brows and reviews the whole dialogue from focus points.
Department in charge
NTT Media Intelligence Laboratories - Social Knowledge Processing Laboratory