WO1999051012A1

WO1999051012A1 - Procedure for establishing a call by means of voice rooting

Info

Publication number: WO1999051012A1
Application number: PCT/SE1999/000506
Authority: WO
Inventors: Lars Devallius
Original assignee: Telia Ab (Publ)
Priority date: 1998-03-28
Filing date: 1999-03-26
Publication date: 1999-10-07
Also published as: NO20004732L; SE520448C2; SE9801125L; EP1070419A1; SE9801125D0; NO20004732D0; EE200000565A

Abstract

Procedure at a telecommunications system, which consists of the public telecommunications network consisting of the fixed network PSTN (1), the mobile network NMT/GSM (2), the Internet (3), and subscriber equipment, which makes possible for one or more subscribers, at a certain operator, to, via telephone equipment with own subscription, or via a telephone equipment with other type of subscription, control, via control functions cooperating in the network, calls to the subscriber's telephone equipment with which one wants to communicate. The control is executed by a subscriber transferring, via the telephone equipment, a number which activates a device for verification of the user's authority, which at confirmed control gives access to a specific speech control server (4) and with this cooperating device (5), including a telephone directory, which allows that a calling subscriber in said telecommunications system gives a name via the voice, corresponding a second subscriber to whom the first subscriber wants to be connected, at which the speech recognition is made in said speech control server (4) which derives information from said telephone directory (5) to set up the call.

Description

PROCEDURE FOR ESTABLISHING A CALL BY MEANS OF VOICE ROOΗNG

TECHNICAL FIELD Procedure at a telecommunications system made up of the public telephone network consisting of the fixed network PSTN and the mobile network NMT/GSM and subscriber equipment, which makes possible for one or more subscribers to, at a certain operator, via telephone equipment with own subscription, or via a telephone equipment with another type of subscription, control, via control functions cooperating in the network, calls to that subscriber's telephone equipment with which one wants to communicate .

PRIOR ART

The public telephone network of today includes, when speaking about ordinary telephony between subscribers, the possibility to reach other subscribers by activating, by the fingers, different keys belonging the telephone. This results in that DTMF-tones are generated and signalling is made via the network, at which a call is set up to the one one wants to speak to.

In certain local telephone systems there are ways to, by means of the voice, route calls to the subscriber one wants to talk with. These systems are then based on that a computer identifies a specific name as belonging to a specific telephone number, at which the call in known way is set up. There also exist local terminals which functions in this way, but with a very limited number of words.

TECHNICAL PROBLEM

The problem of the technology of today is that it in many cases is difficult and troublesome to, on the one hand remember the telephone number to the one one intends to call, and on the other to press the keys under certain circumstances .

Another problem is to establish a technical solution which allows a procedure so that all persons under a subscription will have possibility to use a telephone directory which is common to a specific subscription, and to, by means of the voice, call all persons or companies which are connected to the public telephone network.

TECHNICAL SOLUTION

The technical solution is described in what is specified in patent claim 1.

ADVANTAGES By the present invention the public telephone network is improved and makes it more adapted to the customer by a more flexible construction.

The great advantage of the procedure according the invention is the possibility to considerably reduce keying both in the car, at home and at the job. Further, it is possible, according to the procedure, to utilise the same telephone directory from both the fixed and the mobile network, as well as the Internet telephones, which results in that subscribers need only learn one procedure to make voice calls .

Further advantages :

One need not remember the numbers of all friends and acquaintances .

It will be easier and smoother to make calls, and if one in addition has a physical handicap which makes it difficult to use the keypad, the advantages will be even greater.

Increased safety for children. They need not know daddy's number to the job. - All members of the family can use the directory bound to the subscription. DESCRIPTION OF FIGURES

Figure 1 shows a schematic network configuration, according to the invention, in the form of a block diagram.

DETAILED DESCRIPTION

In Figure 1 is shown a schematic structure of the system solution at the realisation of the procedure according to the invention. The Figure shows the public fixed network PSTN 1 and the public mobile telecommunications network 2 and the Internet 3 , which subscribers via any operator can connect themselves by subscribing. To this network a lot of telecommunications equipment is possible to connect, according to known technology, such as fixed and mobile telephones, computer telephones which allows voice calling via the Internet, and other frequent occurring telephone equipment .

The invention is a general procedure where one by means of the voice routes the telecommunications system to set up a connection to the one or those with whom one wants to carry on a conversation. Further, the invention allows that one via a computer 6 can change certain parameters in the equipment which is located in the network in order to get an optimally functioning voice call.

The voice call can be activated from one's own telephone, mobile telephone or Internet telephone, at which call from the telephone is made to a specific number, for instance 020 -number, and the verification is made via A-number. This means that the subscription and the technical number verification is worked out in such a way that only the user of the subscription can use the specific service which is arranged under a specific 020-number. When the call is made via a type of telephone the subscription of which is not one's own, the verification of the user is made by means of a security code. In this case the verification is made by the user giving a specific code, which has been received at the subscription, by pressing a number of keys belonging to the telephone equipment. Alternatively this is can be made by giving a similar code by means of the voice, by pronouncing a specific combination of words or combination of digits or the like. In these cases one in the network located computer, which is adapted for the purpose, will sense this, and give access to the service.

The fixed and the mobile network co-operates in usual way, and routes calls which are intended for the service which is arranged under a specific 020 -number to a speech control server. The incoming speech signal is transformed in the voice recognising equipment 4, which serves as an interface between the speaking person and a computer 5, containing a telephone directory, to digital signals which are transferred and interpreted by said computer 5 and, depending on what the digital signals represent, a voice message is returned via one in the speech recognition equipment located announcing machine 7 which is based on the information which is derived from said computer 5. At indistinct speech or indistinct questions which the speech recognising equipment does not understand, the announcing machine of the system is connected directly, via direct connection, according known technology, and the announcing machine asks attendant questions, for instance "I didn't understand - please try again", in order to execute the subscriber's request. This is made by refined algorithms which are arranged according to specific syntaxes and grammars . The system is a speaker independent system which can be used directly without any training. The dialogue with the system is performed in a natural way with natural speech. This because they are based on a model of the language which has been created by a lot of people speaking phrases and words which have been recorded. At which the examples have been weighed together to a general language model .

Further, the speech recognising equipment 4 is adjusted to break off the execution of a voice message if the system senses normal speech signals during a reading procedure. This gives possibility to quick adjustment of the subscriber's request.

The personal telephone directory is divided into one section for the company, one section for the public, and one private section.

The public part of the telephone directory consists of specific telephone numbers, such as taxi, hospital etc. The personal or own telephone directory consist of the name one selects for the telephone numbers one wants to have in the telephone directory and this part of the telephone directory one can access via computer equipment . On the occasions one wants to change in one's own telephone directory, one consequently can do this oneself via own computer. The public part of the telephone directory is permanently programmed and is provided via the operator one has for one's subscription.

The company part of the telephone directory consists of the number which are specific for a certain company.

The billing system has by this division of personal and public telephone directory possibility to send invoice on one hand to the companies etc which want to be in the public telephone directory, and on the other to those who subscribe to the service which the invention results in.

The service results in that the subscriber of the function voice calling can create a private or personal voice controlled telephone directory, containing up to about 300 records. When the service is connected to a subscription, the service gives possibility to "hot line function" . By "hot line" is meant that the service always is ready for use at lifting of the handset. It gives a great strength to the service that the user can lift the handset, speak the name of the person he/she wants to call, and be connected there directly. All of that by only lifting the handset. Another alternative to offer the service via the telecommunications network is that the subscriber calls a specific telephone number, for instance a 020-number. This number then can, in a simple way, be programmed on the telephone by known technology, at which the subscriber then presses a key and the service is activated.

After that one has access to a voice directory and can use the service.

In the telephone/voice directory which is included in the computer 5 which is connected to the speech control server, about 300 destinations are included which the user can select to use for voice control . These numbers are selected by the subscriber at his/her own discretion. The use of the telephone directory can be exemplified by a scenario:

1. The user lifts the handset and dials 020-111111.

2. The user says "Bengt Bengtsson" .

3. The service connects the call to "Bengt Bengtsson" .

In the telephone directory there are a number of in advance introduced numbers to some important organisations. These are for example Number Information and Telia's Customer Service Taxi. It is possible to change and add new numbers by simple programming in the service of this type if need should arise. In the voice control function one has possibility to change ones introduced voice controlled numbers oneself. This is made by the user via an ordinary computer and an Internet subscription connecting himself/herself to the speech control server. The user then has possibility to change his/her parameters in the telephone directory via a web- interface .

The invention is of course not limited to the above specified embodiment, but may be subject to modifications within the frame of the idea of the invention illustrated in the following patent claims.

Claims

PATENT CLAIMS

1. Procedure at a telecommunications system which consists of the public telecommunications network consisting of the fixed network PSTN (1) , the mobile network NMT/GSM (2), the Internet (3) and subscriber equipment, which makes possible for one or more subscribers, at a certain operator, to via telephone equipment with own subscription, or via a telephone equipment with other type of subscription, control, via control functions co-operating in the network, calls to that subscriber's telephone equipment with which one wants to communicate, c h a r a c t e r i s e d in that the control is executed by a first subscriber transferring, via the telephone equipment, a number and/or address which activates a device for verification of the user's authority, and which, at passed control, gives access to a specific speech control server (4) and with this cooperating device (5) , including a telephone directory which allows that a calling subscriber in said telecommunications system gives a name via the voice, corresponding a second subscriber to whom the first subscriber wants to be connected, at which the speech recognition is made in said speech control server (4) which derives information from said telephone directory (5) to establish the call.

2. Procedure according to patent claim 1, c h a r a c t e r i s e d in that the content of the telephone directory is divided into three separate activable parts containing associated telephone numbers and names and/or expressions which are activated from said speech control server.

3. Procedure according to patent claim 2, c h a r a c t e r i s e d in that the parts of the telephone directory consist of one part for the own company, and one public part, and one part for private use.

4. Procedure according to patent claim 2, c h a r a c t e r i s e d in that changes in the telephone directory is made via Internet by means of Web- interface and/or via voice commands .

5. Procedure according to patent claim 1, c h a r a c t e r i s e d in that, in the speech recognising equipment 4, an incoming speech signal is transformed to digital signals which are transferred and interpreted by said computer 5 and, depending on what kind of information the digital signals represent, a voice message is returned via one in the speech recognising equipment located announcing machine 7, which is based on the information that is derived from said computer 5, at which, at indistinct speech or indistinct questions which the speech recognising equipment does not understand, direct connection is made of the announcing machine 7 which gives a message of the type "I did not understand - please try again" .

6. Procedure according to patent claim 1, c h a r a c t e r i s e d in that the speech signal is generated from natural human speech.