The development of telecommunications services and the search for new applications for them have completely changed the way we look at the telephone as a tool, and its hitherto basic function, resulting from the name itself (telephone), i.e., transmitting voice over a distance, has become just one of many. To illustrate this trend, it is even said that a phone that was previously used by holding it to the ear is now increasingly held in front of the eyes. It is hard to disagree with this statement when we take into account all such services as SMS, MMS, or Java applications. However, there is a category of telecommunications services that require alternating between holding the telephone handset to the ear and in front of the eyes: these are interactive voice services, known in English as interactive voice response or more often abbreviated as IVR. In short, the functional essence of IVR consists of transmitting voice messages in response to user action. In the basic technological model, voice messages are pre-recorded and stored as files optimized for transmission conditions, and the user’s action consists of selecting the appropriate options from the phone keypad, which is transmitted to the system in the form of DTMF signaling. It should be noted, however, that this is not the only possible solution, as in a more advanced version, voice messages can be generated automatically using text-to-speech algorithms. In turn, user choices can be communicated by voice, and their proper system interpretation is enabled by speech recognition technology. The use of both solutions significantly increases the flexibility of service management while reducing the costs of its development and maintenance, and from the user’s side, it improves its intuitiveness. This is well illustrated by a service launched by one of the GSM operators, the purpose of which is to provide up-to-date information on the situation at border crossings. Users calling a special phone number, after a standard greeting, were asked to say the name of the border crossing they were interested in. The system recognized the name and then, using a speech generator, “read out” the saved current information.
The article comes from the magazine “e-Fakty” 1/2005.