Speech Recognition/Telephone
Automated Speech Recognition (ASR) is now commonplace in the field of telephony. The main objective is to do the preprocessing of a structure support request by (ASR) or replace personell in a call center partially for standard structured workflows, e.g.
- ASR for Identification Number:
- (Question) What is you identification number?
- (Answer) sequence of single digits or the number is recognized by ASR.
- (Application) the corresponding file for handling the client case is loaded on the display of the operator
- (Benefit) workload for asking for the identification number and searching for the file of the client can be automatized for call center staff
ASR becoming more widespread in the field of computer gaming and simulation. In telephony systems, ASR is now being predominantly used in contact centers by integrating it with IVR systems. Despite the high level of integration with word processing in general personal computing, in the field of document production, ASR has not seen the expected increases in use.
The improvement of mobile processor speeds has made speech recognition practical in smartphones. Speech is used mostly as a part of a user interface, for creating predefined or custom speech commands.
Learing Tasks
edit- (Areas of Application) Identify application areas in telephony systems in which ASR can be applied and define areas of application that could create risks for the clients that use it (e.g. medical services)! What kind of validation of recognized speech is required for risk mitigation strategies?
- (OpenSource Telephony Server) Analyze the OpenSource Asterisk for integration of speech recognition? What are the challenges and requirements for the integration of Speech Recognition into an existing telephony infrastructure e.g. of educational services for students as a mixture of hotline and automated services driven by speech recognition.
- (Support Service Levels) Staff answering questions on the phone is a precious resource and people that must wait for answers get frustrated (especially during emergencies - analyze COVID-19 and the demand for advice at phone hotlines). Create a workflow that allows to identify steps in the workflow, that are performed prior to the contact to human staff members on the phone the reduce the workload for the clients/students and decreasing the waiting time for the clients/students until "somethings happens". Are there other use-cases that can support clients without the contact to human staff members on the phone?