Speech Recognition

Speech or voice recognition is the ability of a machine or program to recognize and carry out voice commands or take dictation. In general, speech recognition involves the ability to match a voice pattern against a provided or acquired vocabulary.

Usually, a limited vocabulary is provided with a product and the user can record additional words. More sophisticated software has the ability to accept natural speech (meaning speech as we usually speak it rather than carefully-spoken speech).


APPLICABILITY TO CALL CENTERS

Speech and voice recognition are useful tools in call centers in a number of ways. Not all of these tools and uses are exclusive to the call center environment and could be applied to other locations within a SESA[The SeSa Group was founded in 1973 and operates in the Information and Communication Technology (ICT) market]

They also make use of both single-user speaker-dependant PC-based systems as well as multi-user speaker-independent server-based systems.One of the more generic uses of single-user speech recognition is as an aid to Customer Service Representatives(CSRs) and others to reduce or eliminate repetitive stress injuries that result from Keyboard and mouse use, particularly carpal tunnel syndrome. By replacing an individual’s keyboarding of information with a speech recognition interface,

repetitive stress injury complaints decrease significantly. This has proven to be the case in a number of large call centers, one in particular being that of the L. L. Bean Company. A second single-user application is to enable job positions to be filled by physically handicapped individuals. Both L. L. Bean and the State of Michigan UI Appeals division have used speech recognition applications to enable paraplegic individuals to work at jobs they would otherwise not be able to do. This has been accomplished by replacing the keyboarding of information with speech recognition applications.

The other major area of speech and voice recognition use in call center environments is through the use of multi-user speaker-independent server-based systems. These systems are often used in conjunction with existing or new IVR applications. In either case, they serve to extend the capabilities of the IVR applications. As described previously, speech recognition can automate a manual CSR task such as address changes. It also allows a user to use a previously established password or pass-phrase in place of entering a PIN

via the TouchTonekeys on the telephone. As with change of address information, multi-user speaker-independent server-based speech recognition allows a many applications to be developed and implemented that heretofore required a CSR. While most useful for simple multiple-choice selections, similar to an IVR menuing system, it also enables the entry of free-form text by the caller in a very simple, straightforward, and natural way. Whether this information is used to directly update a system’s databases, or if it is first reviewed by support staff prior to being released for system updating, is a matter of management choice. Certainly, during the early phases of a new application’s introduction it might be both cautious and helpful to review any updates before they are made

Key benefits

Client’s application flexibility is matched by its equally flexible benefits.

IVR scalability and reliability by configuring multiple Client-based speech recognition servers to handle thousands of calls simultaneously — should one server ever go down, Client immediately and systematically routes calls to another server.

Rapid application deployment and central administration using Client’s built-in graphical application generator and pre-built components to create “in demand” speech recognition applications, along with the Client system’s single inherent administrative interface to configure one or more speech recognition servers.

A lower total cost of ownership resulting from Client ’s customizable application-driven architecture, open standards interface and unified administration for inbound as well as outbound IVR processes.

A faster return on investment via the ability to b uild and manage results-oriented outbound campaigns, and to pre-integrate call recording, sales verification and other capabilities that enable your company to realize a higher ROI in months, not years.

Speech recognition has developed very rapidly over the last ten years. This has largely been the result of advances in processing power, which has enabled correspondingly rapid advances in speech recognition algorithms. The speech recognition systems that are on the market today are the embodiments of these new algorithms. What was once the property of the R&D lab and science fiction is now a reality.