HK1095235B

HK1095235B - A method of providing voicemails to a wireless information device

Info

Publication number: HK1095235B
Application number: HK07102742.0A
Authority: HK
Inventors: 丹尼尔．迈克尔．道尔顿
Original assignee: 斯皮沃克斯有限公司
Priority date: 2003-04-22
Filing date: 2004-04-22
Publication date: 2012-02-03

Description

Method for providing voice mail to wireless information equipment

Technical Field

The present invention relates to a method for providing voicemail to a wireless information device. The term 'wireless information device' as used in this patent specification should be construed expansively to cover any type of device that has two-way wireless information capability and includes, but is not limited to, wireless telephones, smart phones, communicators, wireless messaging terminals, personal computers, computers and application specific devices. It comprises devices capable of communicating in any way via any type of network, such as GSM or UMTS, CDMA and WCDMA mobile radio communication, bluetooth, IrDA (infrared data transfer), etc.

Background

Voicemail has a dedicated purpose of: when a person attempts to call a user's phone but the user is unavailable, it is used to store a voice message from the person and then deliver the message to the user at a convenient time. But today's voicemail systems, especially for wireless information devices such as mobile phones, are unable to perform this intelligently. The main reason is the interactive nature from the user's wireless information device to the remote voicemail server: typically, a mobile phone user will call a voicemail server controlled by the network operator (or the mobile phone user will be called by the voicemail server). The voicemail server will generate synthesized speech for announcing the number of messages to the user and then playing back the messages; various options are then spoken by the synthesized speech, such as "reply by 1", "delete by 2", "repeat by 3", etc. This presents a number of problems to the user: first, he may not have the pen and paper at hand to remember any important information; second, he may forget or be unable to hear the options, and thus will not be able to effectively operate the voicemail system.

Due to this inappropriate and arcane interaction, at least 45% of mobile phone users do not use voicemail. Among users who use voicemail, voicemail typically accounts for 30% of the user's talk time and expense. One of the reasons for this high degree of possible surprise is: due to the difficulty of interaction, users often dial in the phone again simply to listen to the details of the key message that were not available at the first listening.

In order to make it easier to obtain voicemail, some efforts have been made: reference may be made, for example, to US 6507643 of Breveon corporation: in this patent, a voice mail is automatically converted to a text message using a speech recognition computer, which is suitable for sending as an email message and can be viewed on a text display device such as a PC or laptop. Reading a written message faster than having to listen to a spoken voicemail; and because it has been transcribed, there is no need to write down important information from the message. However, automated voicemail systems have very limited performance and accuracy; they also blindly transcribe normal pauses in human speech ('hiccup', 'kahn', 'o', etc.). When a person is listening to a person speaking, he can easily filter out these sounds and focus on the substantive communication. Blindly transcribing these pauses into email may make the sender appear to be unintelligible.

Theoretically, automatic speech-to-text conversion can also be configured within the mobile phone itself: reference may be made to the Nokia Short Voice Messaging system (see EP 1248486) in which a user may speak a message to his mobile phone, which locally converts it to text using an automatic speech recognition engine, then packages it and sends it as an SMS message.

An inevitable bias in the field of speech-to-text conversion systems is to improve the accuracy of automatic speech recognition software; however, the current generation of software still needs to be trained to recognize words spoken by a particular person, or is limited to recognizing a very limited vocabulary, and presents great difficulties in context. Training requires the user to read a very extensive test passage and then correct transcription errors introduced by machine transcription. This is a very slow and laborious task. While this may avoid using the small keys of a mobile phone to enter text messages, it does not address the inaccurate and inappropriate transcription problems inherent in conventional automatic speech recognition software.

The task of constructing voice recognition software that can reliably and correctly recognize natural speech that relates to any object, anyone, and speaks at normal speed is a daunting proposition. However, in the field of speech-text systems, this is an important goal. The present invention challenges this conventional concept.

Disclosure of Invention

According to a first aspect, there is provided a method for providing voicemail to a mobile telephone, wherein a caller initiates a voice call to the mobile telephone but the voice call is transferred to a voicemail server, whereupon the caller leaves a voice message on the voicemail server; the method comprises the following steps:

(a) converting the voice message into an audio file format;

(b) sending or streaming the audio file to a speech-to-text transcription system comprising a computer network via a wide area network;

(c) one of the plurality of networked computers plays the voice message to an operator;

(d) the operator intelligently transcribing an original voice message into the computer to generate a transcribed SMS or MMS text message;

(e) sending the SMS or MMS text message to the mobile phone,

wherein the transcribed text message includes a unique identification that links the text message to the voice message stored at the voicemail server to allow the voice message to be played to the mobile phone by a user of the mobile phone selecting an option displayed on the mobile phone that relates to the transcribed text message.

Because operator rather than machine transcription is used, the voicemail can be correctly, intelligently, appropriately, and succinctly transcribed into a text message (e.g., SMS/MMS).

There are many advantages to providing voicemail using this method:

A. record for you

o does not dial in to get the message; instead, they have been transcribed into message format correctly and intelligently (e.g., omitting pauses, repetitions, etc.).

o by giving the user an 'inbox' view of the received voicemail, it is possible to see who the message actually came from before opening and reading it, and to control which is read and when it is stored, forwarded, deleted, replied to, etc. The converted voicemail will have a different icon than the standard text message. Where a number is available, the caller ID is used to populate the text message header with the number so that it appears to come directly from that person. Otherwise, if the number is not available, the recipient will see that it is from 'Spin Vox voicemail'.

O key information at hand-when you are supposed to 'move', no search pen and paper are needed

Most new phones (especially smart phones) are able to parse the text and then provide its useful part for automatic use inside the phone and its applications, thereby saving user duplication intersections-e.g. phone numbers can be used for immediate storage or use, addresses can be hyperlinked and added to contacts, or used for automatic positioning on mapping software, etc.

B. Discrete and convenient

O no harassing calls from voice mail when busy. Instead, the user can see that voice messages arrive at the meeting and decide on their own whether to read them.

C. Can always give you messages

The SMS store & forward capability ensures that delivery is fast once the user phone returns to or within coverage

Selecting a convenient delivery method: SMS, E-mail, fax

D. Transferring any other phone supporting call forwarding, e.g. your office phone (landline), to a text conversion service

O landline & mobile voicemail can now be delivered as text to the user's mobile phone — all in one place, convenient at hand

From any phone, mobile phone or landline, voicemail is accessed because the original voice file of the voicemail is stored on a server, where the server can be accessed by any phone using a password

E. When roaming abroad, voice message delivery is inexpensive

O wherever the user is, they can continue to receive the voice message in text form

Instead of receiving them by email, users can get them all over the world and outside the mobile coverage area

O no roaming charges or expensive callback to get the message

Paying account

There are two options-prepaid or postpaid, which can be via a small bill or credit/debit card on the user's telephone bill, and monthly payments can be directly debited. Indeed, any payment method may be used by the commercial service provider via the third party, and thus even PayPal, which is common in the united states, is becoming available as an effective payment method in europe.

Credit/debit card

The user will be able to sign up for a credit/debit card for automatic monthly payments, including direct debit (UK) and PayPal in the united states.

Small bill

The user will be able to purchase SpinVox credits (e.g.. 10 worth) by a single SMS to reply to a payment, which will confirm their newsletter. Generally, this will be of interest to the pre-paid market. Doing so ingeniously avoids the relatively expensive cost (60% +) of multiple individual tiny transactions each time they use the service, which in the opposite case would make it very expensive, and encourages some commitment to the service from the user.

Drawings

The invention will be described with reference to the accompanying drawings, in which:

FIGS. 1-3 are schematic diagrams of the overall voicemail processing process, starting with voicemail initiation, through voicemail processing, and voicemail delivery, in accordance with the present invention;

FIG. 4 depicts a message notification format (displayed in a message inbox of a mobile phone) for a voicemail transcribed using the method of the present invention;

FIG. 5 depicts a conventional text messaging notification;

FIG. 6 depicts how a voicemail transcribed using the method of the present invention appears as a text message displayed on a mobile phone;

fig. 7 depicts a mobile phone displaying a list of text messages in a message inbox. The transcribed voicemail is present in the list; the inset shows how the display would occur if selected;

FIG. 8 depicts a menu listing of three new functions that may be utilized as options related to transcribing a voicemail;

FIGS. 9A through 9D depict a GUI-based voicemail management application for managing conventional audio voicemails;

FIG. 10 depicts the operation of an application that allows a user to speak a message into their mobile phone and remotely convert it to a text message;

FIG. 11 shows the overall flow of actions at the voicemail server to indicate actions initiated by user input;

FIG. 12 shows the overall flow of actions at the voice message transcriber;

fig. 13 illustrates a screen shot of a web page-based interaction used by a voice message transcriber.

Detailed Description

The invention is implemented as part of a group of mobile phone products by SpinVox ltd, london, england:

1.VoicemailView^TM: voicemail-text system-this gives the subscriber the option to deliver the voicemail to their mobile phone as text (SMS/MMS or equivalent message format) with the option to listen to the original voicemail on the mobile phone. The term 'SMS' refers to a short message service for sending a clear text message to a mobile phone; 'MMS' refers to a multimedia messaging service developed by 3GPP (Third Generation Partnership Project) for sending multimedia communications between mobile phones and other forms of wireless information devices. The term also encompasses any intermediate technology such as EMS (enhanced messaging service) and variants such as premium SMS, as well as any further enhancements and developments of these services.

2.VoicemailManager^TM: new voicemail management applications-which add a GUI (graphical user interface) to the mobile phone; it supplements (or replaces) the existing audio menu system (UI) provided by cellular telephone voicemail systems and integrates the call forwarding features of the phone, greeting controls, and other related controls to provide a single environment (application) for voicemail management on the mobile phone.

3.VoiceMessenger^TM: a speech-to-text system that allows users to speak text messages to their mobile phones, convert them remotely to text, and then transmit without the use of a normally tiresome alphanumeric telephone keypad entry system.

The key to accurately transcribing voice messages into text format (as configured in VoicemailView and VoiceMessenger) is: an operator is used to intelligently perform the actual transcription by lifting messages (not verbose verbatim transcription) rather than utilizing an automatic speech recognition system. The key to the efficient operation of this system is the IT architecture: the IT architecture quickly sends voice files to operators and allows them to quickly listen to these messages and efficiently generate transcriptions, and to send the transcribed messages as text messages.

A.VoicemailView^TMVoicemail-text system

Three solutions for a voicemail-text system for delivery are described:

1. inside the network operator-the system is integrated within the operator's network service (see fig. 1).

2. Outside the network operator-the service company accesses the network operator's voicemail system via the landline telephone and provides external services directed to the end user; see fig. 2, or place its own voice mail system within the service company and deliver its services entirely outside the services of the network operator and thus are independent of the network operator and handset, see fig. 3.

1.1 V. ocemailView: inside of the operating side variant

Referring now to fig. 1, the process is configured as follows:

1 caller leaves a voicemail from the PSTN or mobile telephone network.

2 converting the voicemail into an SMS or MMS file through a voice transcription service: this is not performed by the automatic speech recognition system but by the operator. These operators are more accurate and flexible than automatic speech recognition systems and can intelligently interpret the messages, eliminating unnecessary pauses and repetitions, and thereby generating short, simple and intelligible messages. Appendix II defines the requirements for efficient and concise transcription. Operators tend to be able to shorten messages significantly in order to fit them within the maximum of 160 character existing SMS text messages (otherwise fitting longer messages to multiple SMS messages via standard concatenation); however, for MMS, there is no such maximum limit.

Generate a link to the original voicemail file (unique i/d) — this i/d may simply be a hash of the time/date & caller number

Adding said time & date of the voicemail to the header of the SMS/MMS file

Add Caller number to header of SMS/MMS File

3 sending the message file to the SMS or MMS server for storage.

4 sending the message to the wireless terminal via an SMS or MMS gateway.

The 5-user views and manages 'text' voicemails within an SMS or MMS application, or even within a platform dependent messaging application.

6 the user can request to listen to the original voicemail through the new VoicemailManager application running on the terminal (which provides a GUI interface for all voicemail functions; see b.2): play, Fast Forward (FFW), Rewind (REW), next, erase, store, forward, time/date of message, call back (and any other existing voicemail controls available through audio prompts/menus).

7 when the message is 'read', the reliable delivery of SMS/MMS synchronizes the SMS/MMS storage with the voicemail storage.

A.2 outside of the operator variant; service company providing voice-to-text facility for operator referring now to fig. 2, the process is configured as follows:

1 new subscribers provide their phone number, voicemail PIN number, and other details to the service company. This enables the voicemail retrieval and storage servers to access their voicemail boxes to retrieve messages by polling them periodically or by sending new voicemail notifications to them through the voicemail system inside the operator. There are 2 options for the user to pay (prepaid or postpaid):

1. reply text bill (Small bill)

2. Monthly credit/debit card bill

2 callers left a voicemail from the PSTN or mobile telephone network.

3 the service company voice mail acquire & storage server accesses the voice mailbox & 'listen' message of the subscriber:

use of standard DTMF tone play messages, time of call acquisition, caller number and other data to construct the data necessary for text delivery

Create unique i/d-can be simply a hash of the time/date & caller number

Store voicemails for future playback

4 sending the voicemail audio file to the operator-based voice transcription system and converting it to an SMS or MMS file, and to the 3 rd party SMS/MMS gateway for delivery

Generate a link (unique i/d) to the original voicemail file and embed it as information unknown to the user in the SMS/MMS file

Add the time & date of the voicemail to the header of the SMS/MMS file

Add Caller number to header of SMS/MMS File

The MMS file may contain the original audio file 5 embedded for local play to deliver an SMS or MMS message via the subscriber's network operator

Send messages to wireless terminals via SMS or MMS gateway.

The user views and manages 'text' voicemails within the SMS or MMS application, or even within the platform dependent messaging application.

6 users can dial in their voicemails on the network using a new voicemail management application on the terminal (which provides a GUI; see B.2): play, Fast Forward (FFW), Rewind (REW), next, erase, store, forward, time/date of message, call back (and any other existing voicemail controls available through audio prompts/menus).

7 to hear the original voicemail, the user is connected back to the voicemail storage server of the service company. The unique i/d (which the user cannot know in the SMS/MMS message) gets the correct file to play.

A.3 outside operator: voicemail fully provided by a service company

Referring now to fig. 3, the process is configured as follows:

1 new subscribers provide their phone number and billing details to the service company.

They are now using service companies as their voicemail providers.

2 options:

1. they manually divert calls to their phone to the service company voicemail gateway number

2. Service companies provide over the air (over the air) upgrades to change this behavior

There are 2 options for paying (prepaid or postpaid):

3. reply text bill (Small bill)

4. Monthly credit/debit card bill

2 callers leave voicemails from any telephone, typically the PSTN or mobile telephone network.

3 service company voicemail provides all voicemail functions

1. Storing voicemails for future playback

2. Create unique i/d-can be simply a hash of time/date & caller number

4 sending the voicemail audio file to the operator-based voice transcription system and converted to an SMS or MMS file by the operator, and to the 3 rd party SMS/MMS gateway for delivery

Generate a link (unique i/d) to the original voicemail file and embed it as information that the user cannot know in the SMS/MMS file

Add voicemail time & date to header of SMS/MMS file

Add Caller number to header of SMS/MMS File

The MMS file may contain the original audio file embedded for local playback

Delivering SMS or MMS messages via a subscriber's network operator

Send messages to wireless terminals via SMS or MMS gateway.

6 users can dial in their voicemails on the network using standard IVR controls or a new voicemail management application on the terminal (providing a GUI; see B.2): play, Fast Forward (FFW), Rewind (REW), next, erase, store, forward, time/date of message, call back (and any other existing voicemail controls available through audio prompts/menus).

7 to listen to the original voicemail, the user is connected back to the voicemail storage server of the service company. The unique i/d (which the user cannot know in the SMS/MMS message) gets the correct file for playing.

B. Mobile phone software

In any of the above variants, the mobile phone (or other wireless information device with certain characteristics) will need to be upgraded via OTA (over the air) or otherwise:

b.1 viewing voicemail-text messages

There are two options:

1. no modification of the existing telephony GUI-the SMS is simply processed as another message, the SMS being a transcribed voicemail

2. Modifying the GUI to incorporate new features as shown below:

fig. 4 shows a telephone handset icon that can be used next to an SMS message to indicate that it is a voicemail message in a message inbox. The voicemail transcribed to text is present in a message inbox of the device; it has been sent from the Homer Simpson. Fig. 5 shows how the current SMS text icon looks. Another solution is to add some logic to the voicemail before each header, such as "V: "-thus," V: homer Simpson' will indicate that the transcribed SMS of the voicemail is from Homer Simpson. Additionally, inside the text file of the voicemail message, the time and date of the voicemail should be added (since not all gateways correctly time stamp sent messages) as shown in FIG. 6. Fig. 7 illustrates such operation in a mobile phone environment. The user selects the 'read' option for the highlighted transcribed voicemail (from Daniel Davies); the device displays the SMS in the normal way, but with the data and time added. It is also possible to activate the normal audio based voicemail play function only by pressing and holding a specific key (in this figure, key '1').

When a person opens a standard SMS message, further functions are often easily accessible (e.g. via an options menu in a nokia mobile phone), such as 'Erase (Erase)', 'Reply (Reply)', 'Edit (Edit)', etc. Based on this standard 'options' menu or equivalent, this implementation adds three new functions, as shown in fig. 8:

listen to the original file

Callback

Add contact

We detail these new functions below:

listening to an original file: which allows the user to listen to the original voicemail at the moment and to connect correctly to the original voice file using a unique i/d encoded into the SMS/MMS message.

There are three options:

(i) the user enters a standard voicemail system and follows existing audio prompts to listen to the message.

(ii) The user enters a new voicemail management application shown below at b.2.

In either case, when the call to voicemail is ended, the user returns to the same point in the messaging application to decide what to do with the text/audio version.

(iii) The user embeds the original sound file into an MMS message (or equivalent, such as an email) to be played locally on the terminal.

Back-calling

This operation calls back the message using the caller number recorded with it.

Adding contacts

This operation takes the caller's number and automatically adds it to the user's new contact/address entry to be done with a name or the like.

This is a specific example of mobile phone software that is able to parse text that has been converted from speech and use it intelligently. Other examples include:

(a) extracting the spoken phone number to allow it to be used (to make a call), saved, edited or added to a phonebook;

(b) extracting an email address and allowing it to be used, saved, edited or added to an address book;

(c) extract the physical address and allow it to be used, saved, edited or added to the address book;

(d) extracts the web page address (hyperlink) and allows it to be used, edited, saved or added to an address book or browser favorites.

(e) Extracting meeting time and allowing it to be used, saved, edited and added as an entry to an agenda

(f) Extracting a number and saving it to one of the device applications

(g) Extracts the actual noun and provides the option to search or find it on a web page (WAP or full browser).

The extent to which this operation can be done depends on the intelligence of your cell phone (essentially its parsing power and interoperability with other applications and the pervasive clipboards that typically store such data for other applications). Currently, almost all phones support extracting phone numbers, email addresses, and web page addresses from text messages. When a user reads a message, it is typically available with the underlined content (e.g., a hyperlink or equivalent); the user then simply selects 'options' (as found on nokia phone, or equivalent on a different handset) and 'uses' (as found on nokia phone, or equivalent on a different handset), and then proceeds further with context-dependent options (e.g. with the street address it can provide-look up in address book, navigation, save, etc.) depending on the content type.

B.2VoicemailManager^TM: voicemail management application

This application can be used alone or as part of the VoicemailView voice-SMS/MMS system (or equivalent text delivery system) as described above in b.1.

In accessing and managing normal audio voicemails, the voicemail management application gives the user a GUI (graphical user interface) in addition to the standard audio prompts that are normally received. When a subscriber accesses (fig. 9A) their audio voicemail using their mobile phone, they are first brought into their 'voicemail inbox' and then presented using the controls shown in fig. 9B through 9D.

For programming purposes, these controls will almost exclusively involve standard DTMF tones, which the voicemail system uses as input to the user when he now presses a key on their telephone keypad.

FIG. 9A illustrates a user calling voicemail; fig. 9B shows how a new management application is invoked, which displays all voicemail content of the inbox for the first time (here, 3 new audio calls and 2 stored audio calls). The options menu operates as follows:

items listed on the options menu	Movement of
		All play	Playing all messages in order
All deletes	Providing messages to delete — all New and NoticeWith storage-and delete it all
		All marked as listened to	Moving all new messages into a storage folder

Referring to FIG. 9C, if the user selects the category of audio voicemails he wishes to listen to (i.e., new or stored), a menu list of audio voicemails in that category are shown, each identified by the sender's name (if available) or lack thereof, and the caller number. The transcribed text message has ideally added to it the caller name through the transcription service. This includes informing the user when to turn off the voice-to-text conversion in VoicemailView (i.e., they want clear voice mail) so that they can see the name of the person who left them with voice mail before deciding whether to dial in and listen to it. The user can easily navigate to and select the audio message he wishes to listen to. Once the message is selected, a new voicemail control is displayed on the screen as shown in fig. 9C. Their function is as follows:

voicemail control	Movement of
		1 Erase	Erase current message-return to previous screen. The new or stored folder view is used by the user to select which message to listen to now, orAnd then play the next message.
2 next one	Jump to the next message. At the end of the message, return to the previous screen, the folder view is either new or stored.
		3FFW	The message is fast forwarded while the button is held. At the end of the message, the next message to be listened to is stopped and displayed (new or stored folder view), or at the end of all messages, back to the top level view (new folder view)&Storage folder view
4REW	The message is reversed when the button is held. At the end of the message, the previous message to be listened to is stopped and displayed (new or stored folder view), or at the end of all messages, back to the top view (new or stored folder view)&Storage folder view

During this process, the user is always provided with audible navigation options that are synchronized with the content displayed on the screen in order to make them best in both areas. By using voice recognition based on simple commands, the user can only speak the command they wish to perform, so if the user wishes to play a new message, they will only speak "play", and the VoicemailManager engine will recognize this command and perform only that operation — play the message.

Note that: the correct numbers (keypad numbers) and their associated functions are the same as in existing voice mail systems and are changed by the network operator/voice mail system.

B.3VoiceMessenger^TM: voice-text (SMS/MMS) service

It is preferable for users to want to send messages in text format rather than voice-for example, if they do not want to disturb the recipient, but want to give them a message. It is often very difficult for people to type text using fingers on small alphanumeric keypads. They may also be moving (such as walking), or in a car, or only have one hand available, or not be able to type (such as while driving). The VoiceMessenger^TMVoice-text services address this need.

The user enters their messaging/text application running on their mobile phone, simply selects the message recipient from the address book of their phone or by typing in their number, and then selects the new VoiceMessenger option by pressing and holding the '2' key, as shown in figure 10. The user may also be connected to the service to start with and just speak the number or name to a local (on the mobile phone) or remote speech recognition engine that will retrieve the user through the process.

When connected to the remote VoiceMessenger engine, the user simply speaks his message and the remote VoiceMessenger engine records it and then sends the audio file for conversion to text using the operator based voice transcription system. The text formatted message is then packaged into an SMS/MMS (email or other suitable messaging system) and sent through an SMS/MMS or the like gateway. The user will be given an audible prompt to control the input, listen to the conversion and send the message.

C. Extension

C.1MMS Voice Memo (voice-note) to text

Users with phones that allow MMS operation will be able to send voice memos via MMS, then transcribe them and send them to their desired destination based on the operator's voice transcription service. They can also convert the voicemail and send it to their phone in the MMS format if preferred.

C.2 automatic speech recognition

Which will speed up the processing of incoming voice files and reduce operating costs. The main function would be to automatically detect the spoken telephone number and detect the language in order to send the audio file to the appropriate operator at the transcription site. It is also used to detect names and spoken numbers and addresses from the user's online phone book (see below) and commands for VoicemailManager control.

C.3 Online Address book

There will be two forms of online address book that a user can use when connecting to the SpinVox service simply by saying the name of the person they wish to say.

SpinVox online phonebook — logged in via the user's network, they can add the names and numbers of people they wish to have in their SpinVox online address book.

Synching with their Microsoft Outlook (Express or full edition) or other email/PIM/address book client-doing so enables them to bring all their contacts online and not only speak the name of the recipient, but also determine the type of message they wish to send: SMS, MMS, email, fax, etc.

With the network operator, it is also possible to provide SIM backup functionality and then provide them with SIM

The phonebook to facilitate their calling of names therefrom.

C.4 currently available services (Current)

Using the servers currently available, users can define the mode they wish to use for receiving communications, for example 'meeting' lets users know before a communication that the person they wish to contact is in a meeting and will accept the SMS/MMS or VoiceView text message. Once outside the meeting, the user can change his contact status to 'available' and can be reached by telephone.

Appendix I

SpinVox voicemail IVR structure

A standard voicemail server system with IVR is the basis; the IVR is programmed as a flow chart as shown in fig. 11.

2.VoicemailView

The user's phone will reprogram (during the technical provisioning shown below) the ' 1 ' key (standard voicemail access key) to automatically call the SpinVox voicemail server and let them log in automatically (unique phone number + PIN), bringing it to the top level of the IVR tree.

If at any point the user hangs up, the session terminates with the relevant result. If this occurs during recording, including a dropped line from another mobile caller, then the recording is assumed to be complete and the system proceeds to the transcription phase.

Each transcribed voicemail will contain a unique number (depending on the final IVR tree configuration) starting with saying '4' so that when a user presses and holds '1' to connect to the voicemail server of SpinVox, they only press the unique message i/d (e.g., 403), which will bring them into the queue with the third message they own.

2.1 landline or other mobile telephone access

As shown in fig. 11, the IVR tree will allow users to dial in using their unique transfer number (voice mail number) and will then be prompted to enter their PIN.

2.2 speed Dial

The IVR system will accept user programming at speed dial, which allows them to dial their unique SpinVox number + PIN. They can then access all the features shown above.

2.3 leaving voicemail

The user's phone is configured to transfer to SpinVox voicemail under conditions where they are defined as follows, where the caller will hear one of the following:

default SpinVox greeting: "welcome to SpinVox voicemail. Please clearly dictate your message after beeping. "[ Dudu]

User's own greeting: [ recorded greeting of user][ Du]

Then:

1. the system records the caller's voicemail in a default length (30 seconds) or a user-defined length (10 seconds-2 minutes or any SpinVox set parameter).

2. At the end of the recording, the caller listens to standard IVR options by prompting: "press:

1. listen to your message

2. Delete your message and re-record

3. Re-record your message

End or hang up only "

3. If the user exceeds the record length, they are prompted to: "wrong, you have exceeded the available recording time. Please retry after beep ".

a. If the user hangs up without recording a new message, a message is sent for transcription.

b. If the user has selected 'advanced transcription option', another variant appears; this option operates so that if the recording time of the message is less than the maximum time set by the user, the message is transcribed, otherwise, it is not transcribed, but a standard notification is sent to the user to inform them that there is a new voicemail to listen to in accordance with the format shown in 4c below. This operation illustrates the fact that: occasionally, long voicemails are sent to the user that are more favorable to listen rather than read. For these long messages, however, the transcriber may briefly listen to the voice message and write a very short indication of the subject of the call that is sent to the message recipient. Furthermore, for handsets that support less than a certain amount of text (typically legacy handsets), the system first looks up the user's handset and the restrictions in the phone database (provided by SpinVox) and then provides the user with the associated record length. For example, for an old siemens phone that does not support links and only has 4 text messages at most, the system alerts the user to: the record length should be kept below say 30 seconds to ensure that most messages fit on their phone and tell them why. Also, the system is required to set the default record length for these handsets to the appropriate length for them.

4. Sending messages to the relevant transcription queue:

a. if Caller CLID (calling Line Identification) is acquired, the 'From' field is automatically filled in. If not, insert 'SpinVoxVoicemailView' as the sender.

b. If transcription is possible, the text version of the message is sent to the user

c. If transcription is not possible, then a template text message with certain fields auto-populated is sent to the user:

"you have [ from CLI, if available ] a new voicemail to listen to. Connect to your voicemail as '1' on your phone and then listen to this specific message as per 4 xx. Thanks to the culture. SpinVox ". The 'From' field is From 'SpinVox VoicemailView'

d. Payment is made in accordance with the number of SMS transmissions.

5. Text messages are sent to the user and they can choose what to do next according to standard options available on their handset.

3.VoiceMessenger

The IVR chart above shows how users access VoiceMessenger directly from their mobile phone or through another phone.

3.1 speed Dial

The IVR system will accept user programming at the time of speed dialing, which allows them to dial their unique SpinVox number + PIN + '3'.

If their mobile phone is utilized, the following technique provides that a speed dial (via default key '2') will be configured to dial and have them directly log in (voicemail number + PIN +3) to the VoiceMessenger option.

They will then hear the standard prompt:

"welcome to VoiceMessenger of SpinVox. After beeping, please speak the destination number or type it in, and then dictate the message you wish to send. Hang up to send, or send new messages by #. "[ Dudu ]

Then:

1. if DTMF tones cannot be detected, or confusion occurs (when international dialing is done using star or +), then prompt for a new number:

"wrong up, we cannot detect the number you type. Please retry and remember that the international number starts with 00 without + "[ beep prompt re-entry ]

2. The system records within a default length (30 seconds) or a user-defined length (10 seconds-2 minutes).

3. At the end of the recording, the user hears the standard IVR option by prompting:

"press:

4. listen to your message

5. Delete your message and re-record

6. Re-record your message

Sending new message or hanging up only "

4. If the user exceeds the record length, they are prompted to: "wrong, you have exceeded the available recording time. Please retry after beep ".

5. Send message to the transcription queue, where the 'From' field is automatically filled (when SpinVox knows who is the client):

a. if transcription is possible, the text version of the message is sent to the user

b. If transcription is not possible, a template text message is sent to the user that automatically fills in certain fields:

"sorgo, but we cannot convert the message [ time/date ] you dictate [ to number, if detected ]. Please retry in a quiet environment and clearly dictate. Thanks to the culture. SpinVox ". The 'From' field is 'SpinVox VoiceMessenger'.

c. Payment is made according to the number of SMS sent or MMS size (kilobyte).

6. Text messages are sent to the recipient and they can choose what to do next according to standard options available on their handset.

4. Technical provision

During the technical provisioning, the user data (handset, network, etc.) will be reused to confirm what the user has selected.

It is critical that the system sending the user SMS message automate the configuration part of the user handset (transfer & v.card for VoiceMessenger) and confirm the successful setup. These messages are all sent as high priority to ensure that the user/salesperson does not 'hang up' (hang) while waiting for the configuration SMS to succeed.

The steps are as follows:

step 1: mobile phone selection (typically at the point of sale) from a drop-down list displayed on a provide screen

Step 2: voicemail view settings:

< creation of the following symbol string: ' + COUNTRY CODE _ USERS UNIQUE

VOICEMAIL NUMBER _ p _ PIN NUMBER _ # - > > > > this is called the SPINVOX VOICEMAIL NUMBER and is unique to each user! < CHEM > A

And step 3: and (3) call forwarding selection: which explains how a mobile phone is typically set up to transfer to the user's voicemail (based on all of the following). The user can change these if he specifically wishes to transfer to another person or number instead of his own voicemail

The USSD symbol string … (digit line) created based on the above selection for configuring the handset is sent as a high priority SMS with a 4 xsussd symbol string that the user needs to answer/act. < CHEM > A

And 4, step 4: call forwarding setup via SMS. Telling the customer that he has just sent an SMS and should click on a designated button on the offer screen when received (or additionally click on the 'not received' button if not received within 3 minutes).

And 5: and setting call forwarding: SMS. The provide screen informs the user if he has received the configuration SMS, please perform the following steps:

1. opening SMS messages

2. Select 'Option' (database with function name for each handset)

3. Scroll & select 'used number'

4. You will now see 4 numbers, select the first number and press send.

You will now see the number being dialed and display 'requesting' on the screen of your mobile phone. If you receive the confirmation message, this step is repeated for the remaining 3 numbers.

And 5: and setting call forwarding: a mobile phone. The provide screen notifies the user:

on your mobile phone:

1. select 'menu' 2.

< import VOICEMAILVEW data … from database for specified handset tells you what to do/utilize '+ COUNTRY CODE _ USERS UNIQUEE VOICEMAILVANILNUMBER _ p _ PIN NUMBER _ #' >)

Step 6: a delivery method is selected. Once the voicemail is converted to text (typical options are SMS, MMS, email with audio files), the screen is provided to allow the user to select how he wants to receive the voicemail. The system then sends the appropriate vCard to the user's mobile phone.

And 7: a voice messenger setting. The provide screen notifies the user:

please perform as follows:

we just send you an SMS-VCard. When you have received it, please perform the following steps:

1. accept and save the VCard on your mobile phone without modifying it-go to step 2.

If you do not receive this message within 5 minutes, or if you cannot save the VCard, then please perform the following steps:

a new 'contact' called 'VoiceMessenger' is created, with the following number:

+COUNTRY CODE_USERS UNIQUE VOICEMAILNUMBER_p_PIN NUMBER_#_，1′

if you don't know how to add a new ' contact ', please click here- (go to the ' how ' page, which has information obtained from the database-telling you what to do)

< import VOICEMESSENGER speed dialing configuration data … from database for specified handset tells you what to do/utilize >

And 8: congratulation screen:

thank you to select the SpinVox service.

You will receive the voicemail as text at this moment and not forget: you can always listen to the original mail just by pressing and holding the '1' key on your phone (to connect with your SpinVox voicemail account).

To speak a text message-press and hold '2' (or you designate a key of VoicemailView) and you will be immediately connected to VoiceMessenger. Clearly dictating your number and message-you say it … we convert it to text!

You can always access VoiceMessenger by pressing and holding the '1' key and following the prompt.

You can use your mobile phone number and PIN to view your account settings, view accounts and manage your SpinVox account at www.SpinVox.com.

If you have not printed or recorded your PIN number, get it again here

1234

5. Transcription assist

When the operational transcriber registers their account, this is provided to them. All that they need is a web browser, sound card, media player capable of playing and controlling media file playing or streaming protocols, and high speed internet access. FIG. 12 shows a process flow diagram for transcription. Each transcriber logs in and starts receiving VoicemailView (see fig. 13, the screen in which they type the transcription message and are used to cause the message to be sent), or VoiceMessenger audio files to be transcribed (see fig. 14), one at a time. When logging in, there are only 2 states: the message is currently in the process of being transcribed, and is paused.

5.1 transcriber control panel button (see FIG. 13):

completed transcription

Untranslated transcription-according to above 2&3：

For VoicemailView an automated SMS with an auto-filled field (if available) is sent to them as follows, the text is "you have 'From CLI', if available ] a new voicemail to listen to. Connect to your voicemail by '1' on your phone and then listen to this specified message by 4 xx. Thanks to the culture. SpinVox. "

The 'From' field is From 'SpinVox VoicemailView'

To VoiceMessenger, when data is available, an automated SMS is sent to them with an auto-fill field according to the following, the text being: "don't care, but we cannot convert your time/date" to phone number if you get the spoken message. Please retry and clearly dictate in a quiet environment. Thanks to the culture. SpinVox ".

O the 'From' field is 'SpinVox VoiceMessenger'.

Pausing and re-queuing current information

Rerouting current information to a different language, menus for selecting languages or

"unknown". The transcriber is returned to the queue to receive the new message.

5.2 telephone number:

in the case of VoicemailView, the 'From' field is utilized when the caller leaves informationTaken CLID (inserted into message header) or "SpinVoxVoicemailView" automatically filled

In the case of VoiceMessenger, the 'From' field is automatically populated for the transcriber when DTMF is used by the user, or if DTMF is not used, the transcription assistance provides a field to the transcriber to enter it.

Note that: for user data protection reasons, the transcriber never sees the auto-populated telephone fields (or other user data fields), so the system will not show these unless it requires the transcriber to enter a destination number.

5.3 spell checking

When the transcriber taps 'send', the system will automatically spell check the message and if any errors occur, correct them and display the corrected results to the transcriber with the prompt "accept & send" or allow them to manually correct (as there may be a special spelling they wish).

To perform this correctly, the spell-checking process will include a dictionary of real nouns that are relevant to the user's geographic area and culture. Thus, for example, in the uk, a real noun dictionary will contain not only english names, but also place names, landmarks, road names, chain store names (e.g., pubs, bars, restaurants, etc.), and the like.

When there is no match, the transcriber simply double-clicks the underlined word to provide the closest match. If necessary, they can rewind and listen back to the portion of the message to make the appropriate selection.

5.4 transcriptional department Administrator

They can view statistics for all of the transcriber accounts in their own name. They will be able to view and analyze:

number of transcripts entered hourly, daily, weekly, monthly, yearly (sign on, support)

Number of SMS transmissions entered hourly, daily, weekly, monthly, yearly

Queue time per hour, day, week, month, year

Average message length entered hourly, daily, weekly, monthly, yearly

Hourly, daily, weekly, monthly, yearly number/rate of transcriptions

Change in number/rate of transcriptions entered hourly, daily, weekly, monthly, yearly

Post of transcribing party accountThere are these

The number and% of messages that cannot be transcribed entered daily, weekly, monthly, yearly

The number and% of messages sent to different locations for transcription per day, week, month, year

Transcription accuracy-by daily random sampling and measuring accuracy relative to the original (CCA manager performs this operation)&Entering the results into the system) and feedback from the CCA regarding the incident ticket. The worse of these two numbers is accuracy.

Appendix II: requirement for transcription service

These are requirements for transcription services for VoicemailView and VoiceMessenger services.

Require that

The key requirement is to deliver the actual message, rather than all redundant information that is often spoken and left in the message.

Require that

Privacy transcription services must be minimized within the data protection act 98 or other act in effect at the timeTo provide complete confidentiality of the transcribed message. All transcription employees must sign a secret agreement before being able to process any messages and not leak, share, copy, forward or otherwise share any user information-separating messages and numbers to protect user information: in the case of VoicemailView, the transcriber does not show the user phone number to which they are sending a text message-in the case of VoiceMessenger, they do not see the caller's number, only the destination number

09]

[0310]

For example, speaking a message: "you are seen outside of the wax O 'Connors, where wax is wax in candle wax and O' Connor is O 'Connor of the irish singer, Sinead O' Connor. "transcription should read: "see you outside of wax O' Connors. "
	Abbreviations
Standard abbreviations for general terms should be used: the abbreviation Apartment Number Telephone Number Tel. Fax Number Example (e.g. E.g. good) ok Electronic Mail email Intemet Website) Website (i.e. without http: /)
	Number ofWhenever a digit is spoken, the digit format will be written down. For example, "see you at seven forty-five times a night" see you at 7:45pm "e.g.," we want to order ten thousand seven hundred eighty-eight D4 sized nuts. By "═ we intended to purchase 11,788 nuts of size D4. "for example," Jane lives in the eleven-seventy-five Park landscape (Park View), twenty-three rooms on the fourth level "— Jane lives in 1175 Park landscape (Park View), 4^rdLayer apt.23. "
Telephone number to save character space, a telephone number is a single string of numbers without spaces: for example: 07798625155 instead of 07798625155 as with the use of two additional space characters.
	International prefix if a telephone number is given in 00 when international dialing is performed, it is converted to '+'.

[0311]

For example, 00442075864103 should be + 442075864103. Doing so further saves character space and correctly defines the number of the international dialing prefix, which is translated by the home network into the appropriate international dialing code, which is not always 00 (e.g., it is 011 in the united states).
	Spell-check messages must be spelled correctly and it has been suggested that a related spell-check program be used for all messages-e.g., english for the united kingdom, english for the united states, etc.
The dictionary/spell checker used by the actual nouns and place names must include the actual nouns (names) and place names to help get the information in the message correctly for the first time.
	Event plan-a daily calendar of events, celebrations, news, etc. There are several aspects to it: (i) cultural proverb in a multicultural society, it is important to know that something will be celebrated in a certain group on many days. For example, india (Hindi) new year (Divali) is not the same as the main UK new year, so for Divali the transcriber must prepare to listen to greetings and blessings with these and other associated words in them, and know how to spell them, or know what the speech of the message might mean. (ii) normal annual events-Easter, Christmas, New year, etc. (iii) sporting events-national tournaments, world cups, F1 events, sailing events, and the like. (iv) media events-awards of Oscar, English movies, and television arts college prizes (BAFTA), etc. (v) eventualities-like recent 'world trade center double tower' attacks, Madri explosions, Iraq wars, etc. The local transcription department administrator must have a full calendar of all cultural, social, sports events, for which he must plan at least 2 days in advance. In addition, this will be the key to determining the possible burden balance required by the personnel. For example, at the end of the last uk football world cup win, the text message and voice mail loads are probably above 300% of their normal levels within 2-3 hours after the race, and often mostly relate to player names, terms used in the race (try, conversion, rake, hill, etc.), foreign city cities (e.g., football, etc.), and the like

[0312]

And location, of course, on subsequent days, all traffic associated with people returning from the event will naturally bias the burden balance again.
	The untranslated word, after making the best effort to find out what this word is (perhaps the name of a bar or place outside the usual vocabulary), will then have a question mark placed in parentheses. For example, speaking a message: this 6 o' clock meets you at Jongleurs. Transcription: this evening 6 o' clock meets you at Junglers (.
Gap or drop messages (drop-out) may contain 'drop', 'gap' or other disturbances due to the occurrence of temporary network failures. In this case, "____" is inserted where a word is missing. For example, "John, Mike and I ___ arrive ___ late, so you see you at 6 pm. This would likely prompt the user to dial in to listen to the original file and see if they can understand the message. And (3) disconnection for more than 3 times: in this case, since the number of times a word is dropped is too many (3 times or more), the message is difficult to understand, and then an 'untranslated' option is used to send a notification to the user telling them that they need to listen to a voice mail or try to speak their text message again.
	An untranslated voice message will inform the user through a text message using a standard template: for their listening, there are untranslated voice messages: VoicemailView the standard text would say "you have x new voicemails that cannot be converted to need to be listened to. To listen to them, please connect to VoicemailView by holding and pressing 1. "then, the following fields will be automatically filled: caller [ telephone number ] when CLI is disabled]Or [ private number ] "]Time/date][ unique i/d]So that the user can directly obtain the message

[0313]

The standard text described by VoiceMessenger would say "sorry, we cannot convert the message you just spoken. The retry is somewhat slower and clearly spoken. Thanks! "then, the following fields will be automatically filled: time and date]They attempt to send a message to: [ telephone number]They are trying to convert to text
	Tone or other implied context when it is apparent to the person leaving the message that the tone was also used as part of the message, then the transcriber will include the following at the beginning of the message: chinese character of' Xiao]Laugh [ crying]Cry [ whisper]Whisper [ yelling]Shouting/screaming (unless this is done to overcome background noise when in a bar or the like, and if so neglecting) · screaming]Screaming sound when someone is very painful, criticized or scared. [ frightening terror]When a person becomes terrorism [. vital essence generation ]]This is not added when the vitality of shouting and/or punching a fist (which should emerge from the content of the message) is unclear of tone (which may be exactly the way the person is talking or the voice they are in, for example).
It has become very common for VoiceMessenger text style (textism) to insert text symbols to represent emotions (emoticons). The following will be published and will be supported. These are collections that we will support and publish on our websites. The official full list of SMS-Speak is:http://sites.ninemsn.com.au/minisite/web2sms/help/smsdict.asp during dictation of a VoiceMessenger message, a user may speak a VoiceMessenger messageSay "insert symbol name", the transcriber will insert the appropriate symbol. For example, "thank you confirm our journey. A smile is inserted. Turning to! "thank you confirm our journey: -) see again! "

[0314]

Claims

1. A method for providing voicemail to a mobile phone, wherein a caller initiates a voice call to the mobile phone but the voice call is transferred to a voicemail server, whereupon the caller leaves a voice message on the voicemail server; the method comprises the following steps:

(a) converting the voice message into an audio file format;

(e) sending the SMS or MMS text message to the mobile phone,

2. The method of claim 1 wherein the transcribed text message has added a time and date that the voice message was originally received at the voicemail server.

3. The method of claim 1, wherein the voice message is initiated at a mobile phone or a landline phone.

4. The method of claim 1, wherein the transcribed text message has added caller identification data.

5. The method of claim 4, wherein the caller identification data is a telephone number.

6. The method of claim 4, wherein the caller identification data is a caller name.

7. The method of claim 4, wherein the transcribed text message is displayed on the mobile phone.

8. The method of claim 5, wherein the computer does not display the phone number associated with the mobile phone to the operator.

9. The method of claim 1, wherein the computer displays an option to the operator to reroute the audio file to a different computer that makes the operator more appropriate to transcribe the voice message due to linguistic, dialect, or cultural reasons.

10. The method of claim 1, wherein the computer provides the operator with a searchable list of terms related to cultural proverbals, regular events, media events, and other types of newsworthy events to assist the operator in transcribing those terms correctly.

11. The method of claim 1, wherein the operator indicates in the transcribed text message an emotion of the caller who left the voice message using a written description or an emoticon.

12. The method of claim 1, wherein the operator succinctly summarizes the voice message.

13. The method of claim 11, wherein the operator summarizes the voice message to conform to a 160 character SMS limit.

14. The method of claim 1, wherein the operator ignores from the transcribed text message any pauses, artifacts, or repetitions present in the voice message.

15. The method of claim 1, wherein the text message is sent to the mobile phone in a format previously specified as appropriate by a user of the mobile phone.

16. The method of claim 1, wherein the text message is sent as an SMS, MMS, email, or fax.

17. The method of claim 5, further comprising the steps of: parsing the transcribed text message and using the parsed data in an application running on the mobile phone.

18. The method of claim 17, wherein the step of parsing the transcribed text message and using the parsed data involves one or more of the following steps:

(f) extracting said telephone number to allow it to make a call, to be saved, to be edited or to be added to a phonebook;

(g) extracting an email address and allowing it to be used, saved, edited or added to an address book;

(h) extract the physical address and allow it to be used, saved, edited or added to the address book;

(i) extract the web page address and allow it to be used, edited, saved or added to an address book or browser favorites;

(j) extracting the meeting time and allowing it to be used, saved, edited and added as an entry to the agenda;

(k) extracts the actual nouns and provides options to search or find them on the web page.

19. The method of claim 1, wherein a voice message is transcribed if its recording time is less than a maximum time set by a user of the mobile phone, otherwise, no transcription is performed and a notification is sent to the user of the mobile phone that they have a new voicemail to listen to.

20. The method of claim 19, wherein an operator listens to the voice message and a written record is sent to a message recipient of a very short indication of the subject matter of the call.

21. The method of claim 19, wherein for mobile phones that support less than a certain amount of text, text limits are first looked up in a database, followed by automatically suggesting a suitable maximum time.