Data processing: speech signal processing – linguistics – language – Speech signal processing – Synthesis
Reexamination Certificate
1999-12-20
2004-01-13
McFadden, Susan (Department: 2655)
Data processing: speech signal processing, linguistics, language
Speech signal processing
Synthesis
C704S270000, C704S270100, C704S275000, C379S088010, C379S088130
Reexamination Certificate
active
06678659
ABSTRACT:
The present invention concerns a system and a method of dissemination of voice information. More specifically, the present invention concerns a voice teleservices system and method, allowing a user to access by means of a telephone, or any type of device equipped with acoustical emission means, information disseminated by a centralized platform in a telecommunications network.
In spite of the emergence of other systems of automatic dissemination of information, for example Internet or teletext, the offer of voice teleservices tends to grow. These services can in fact be used without special equipment, except for a telephone handset, and can thus be easily consulted from anywhere, for example by means of a portable telephone. Moreover, access to the information is possible without particular technical knowledge. Examples of popular voice teleservices include, for example, the talking clock, dissemination of weather forecasts, sports results, stock market prices, etc.
The information disseminated is generally stored in digital form or registered on an audio data recording medium in a centralized platform connected to the telephone network. The user takes the initiative to consult this information by dialing on his telephone handset the telephone number of the platform. An ordinary telephone connection is then established between the centralized platform and the user, and a vocal dialogue is carried out via this connection. The connection is bidirectional, even when the information is disseminated in a sole direction (this is typically the case for a talking clock service), or primarily in one direction (for example in the systems where the only expected responses of the user are limited to the introduction of a password or short-responses of the yes-no type). The load caused on the telecommunications network is thus out of proportion with the actual volume of useful information transmitted.
The progress observed for some years in the field of voice processing, in particular of voice synthesis (TTS systems—“text-to-speech”), of voice analysis and of personal voice recognition, opens the door very wide to new systems of voice teleservices, extending the possibilities offered by the conventional systems. The field of application of these new technologies is no longer limited to unidirectional information dissemination services (such as the talking clock), but also relate to interactive information services, offering possibilities of dialogue between the caller and the voice teleservice platform. These interactive systems are generally known under the anglophone initials IVR (Interactive Voice Response Systems), and are applied in particular to telebanking services. Already known are systems offering the user the possibility to choose, by means of voice commands, the information he wishes to hear, or even change this data, or initiate the launching of application programs executed by the voice teleservice platform. By way of example, the patent document WO88/05239 describes a system allowing polls or voting to be carried out automatically. WO93/26113 describes another largely automatic system of voice messaging.
The voice teleservice systems are generally operated by means of a centralized platform, for clear reasons of updating information and of cost. The length of the telephone connection established thus depends upon the distance between the calling point of the user and the location of the platform. The attractiveness of voice teleservice diminishes greatly, however, for the subscribers who cannot call at local rate. In numerous cases, the management of the telecommunications network therefore proposes a single calling rate for a given teleservice, which penalizes the subscribers closest by and shows a deficit for the calls of subscribers residing farther away.
The patent document EP-A2-0 559 981 describes an interactive voice teleservice system in which the user is connected to the centralized platform by means of a completely numerical connection. The object of this system is essentially to limit the number of analog-to-digital and digital-to-analog conversations between the user and the voice teleservice platform. The problems mentioned above are not resolved by this system, however.
An object of the present invention is therefore to propose an improved voice teleservice system. More specifically, an object of the present invention is to propose a voice teleservice system resulting in a lesser load on the telecommunications network.
These objects are attained in particular by means of the elements of the characterizing parts of the independent claims, the preferred embodiments being indicated in the dependent claims.
The invention has as its point of departure the observation that, in the prior art, the voice messages exchanged, in particular the voice information disseminated by the platform to the caller, require a bandwidth much larger than the minimum required to transmit the sole semantic content of the transmitted information.
The objects of the invention are therefore attained by transmitting, not analog or digital (phonetic) audio signals, but voice messages containing only one piece of information on the semantics of these signals.
In the case of a digital system, the transmission of thousands of successive voice samples, coded, for example, on 8 bits, is thus replaced by the transmission of some characters, for example some ASCII codes corresponding to the representation in text mode or pseudo-text mode of the semantic content of the messages.
The compression rate thus obtained, and therefore the load on the network is very high, typically on the order of 1:500, for example, depending upon the messages and the type of application. The conversion of semantic message into audio signal is carried out preferably by a network access point close to the user, so that an audio signal of large bandwidth is transmitted solely between the network access point of the user, that is, on a fraction of the total distance separating the user from the teleservice platform.
In view of the very low rate of information obtained by means of the method according to the invention, and the irregular nature of this rate, in particular in the case of interactive systems, the coded messages in semantic form are preferably transmitted in the form of packets between the centralized platform and the network access point or points. The telecommunications network preferably connecting the centralized platform to the network access points is therefore made up of a packet transmission network, for example an ATM protocol network and/or Internet (TCP/IP). In the latter case, the network access points can preferably be made up of distributed POPs (Points of Presence) of the Internet network.
The user wishing to call on a voice teleservice system, for example to listen to weather forecasts, calls the closest POP by means of his telephone. This connection can generally be established at local rate. The POP then contacts the centralized platform managing the automatic voice teleservice system, which transmits to him a coded message, including a semantic representation, containing the required information, for example a textual file announcing the weather forecasts for the region under consideration. This textual file is converted into audio signal (analog or digital) by means of a voice synthesis device in the POP (TTS device “text-to-speech”), and the audio signal obtained is transmitted to the user, who can then listen to it.
An ordinary telephone connection is required only between the user and the network access point (POP). Considering the great concentration of Internet network access points, this connection can very often be established at the local tariff. Between the POP and the platform, the connection is preferably of Internet type, and its cost therefore does not depend upon the distance. Moreover, the volume of information exchanged between the POP and the platform is very much reduced, owing to the conversion in semantic form of voice messages.
In the case where a short response time is ne
McFadden Susan
Oblon & Spivak, McClelland, Maier & Neustadt P.C.
Swisscom AG
LandOfFree
System and method of voice information dissemination over a... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System and method of voice information dissemination over a..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method of voice information dissemination over a... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3186580