Voice site personality setting

• US 7,016,848 B2
• Filed: 12/04/2001
• Issued: 03/21/2006
• Est. Priority Date: 12/02/2000
• Status: Expired due to Term

• Current Assignee: ShoreTel Incorporated (Mitel Networks Corporation)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Squibbs, Robert Francis, St John Brittan, Paul
• Primary Examiner(s): ABEBE, DANIEL DEMELASH
• Application Number:
  US10/005,904
• Publication Number:
  US 20020099553A1
• Time in Patent Office: 1,568 Days
• Field of Search: 704/270, 704/275, 704/235, 704/260, 379/88.03
• US Class Current: 704/275
• CPC Class Codes: G10L 13/08 Text analysis or generation of parameters for speech synthesis out of text, e.g. grapheme to phoneme translation, prosody generation or stress or intonation determination
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications

Abstract

A method is provided of setting the voice personality of a voice service site. A user browsing a voice web visits a voice site where the voice output of the site is presented using a set of voice personality characterisers with which the user is particularly comfortable. The user, in subsequently transferring to another voice service site, opts to have the voice personality that was embodied in the set of voice personality characterisers used by the site being left, transfer with the user to the new site. This transfer will typically be subject to permissions set by both the site being left and the site about to be visited.

Claims

1. A method of setting the voice personality of voice service sites that operate by serving, from voice page servers, voice pages in the form of text with embedded voice markup tags to a voice browser, the voice browser carrying out text to speech conversion to generate voice output, and the voice browser being disposed between voice page servers and the user;

wherein a set of voice personality characterisers identified by indications included in a voice page of a previously-visited voice service site and used by the voice browser for presenting voice output of that page, is also used by the voice browser in presenting the voice output of a voice page of a next-visited voice service site.

2. A method according to claim 1, wherein text to speech conversion at the voice browser is effected using a voice tract model, said voice personality characterisers being parameter values of this model.

3. A method according to claim 1, wherein text to speech conversion at the voice browser is effected by the concatenation of speech elements such as phonemes, allophones, diphones or the like, these latter forming said voice personality characterisers.

4. A method according to claim 1, wherein a user accessing said voice service sites is asked at the time of transition from the previously-visited voice site to the next-visited voice site whether the set of voice personality characterisers used in presenting the output of the voice page of the previously-visited site is to be used in presenting the voice page of the next-visited site.

5. A method according to claim 1, wherein phrasing style data is associated with said set of voice personality characterisers, the voice browser using this phrasing style data to generate text, for subsequent conversion to voice output, that expresses a specified semantic meaning in a particular style matched with the voice personality defined by the voice personality characterisers.

6. Voice browser apparatus for handling voice pages comprising text with embedded voice markup tags, said apparatus comprising:

voice-service access means for accessing voice pages of different voice service sites;

a text-to-speech converter for presenting the output from an accessed voice service site using a voice set by a selected set of voice personality characterisers; and

selection means for selecting the set of voice personality characterisers to be used by the voice-service presentation means to present the currently accessed site, the selection means being operative to enable a set of voice personality characterisers identified by indications included in a voice pane of a most recent previously-accessed voice service site and used by the to-speech converter to present the voice output of that page, to be selected for use in presenting the voice output of a voice page of a currently-accessed voice service site.

7. Apparatus according to claim 6, wherein the text-to-speech converter is of a type using a voice tract model, said voice personality characterisers being parameter values of this model.

8. Apparatus according to claim 6, wherein the text-to-speech converter is of a type that uses concatenation of speech elements such as phonemes, allophones, diphones or the like, these latter forming said voice personality characterisers.

9. Apparatus according to claim 6, further comprising means for checking with a user at the time of transition from said previously-accessed voice site to said currently-accessed voice site whether the set of voice personality characterisers used in presenting the output of the voice page of said previously-accessed site is to be re-used in presenting the voice page of the currently-accessed site.

10. Apparatus according to claim 6, wherein the voice browser includes a language generator arranged to use phrasing style data associated with the selected set of voice personality characterisers to generate text, for subsequent conversion to voice output by the text-to-speech converter, that expresses a specified semantic meaning in a particular style associated matched with a voice personality defined by the selected set of voice personality characterisers.