Method for converting a text classified ad to a natural sounding audio ad

• US 5,832,432 A
• Filed: 01/09/1996
• Issued: 11/03/1998
• Est. Priority Date: 01/09/1996
• Status: Expired due to Term

• Current Assignee: Qwest Communications International Incorporated (Lumen Technologies, Inc.)
• Original Assignee: MediaOne Group Inc. (Comcast Corporation), US West, Inc. (Lumen Technologies, Inc.)
• Inventors: Erbes, Carol A., Warden, Barbara J., Case, Eliot M., Trader, Terry F., Bixler, Matthew J., Bishara, Michael J.
• Primary Examiner(s): Dorvil, Richemond
• Application Number:
  US08/584,414
• Time in Patent Office: 1,029 Days
• Field of Search: 395/2.67, 395/2.69, 395/2.79, 395/2.81, 395/2.86, 395/214, 395/226, 395/227, 395/796, 395/792, 395/793, 704/260, 704/272, 704/275, 704/270, 704/277, 704/278, 704/258
• US Class Current: 704/260
• CPC Class Codes: G10L 13/027 Concept to speech synthesisers; Generation of natural phrases from machine-based concepts generation of parameters for speech synthesis out of text G10L13/08
• CPC Class Codes: H04M 2201/60 Medium conversion
• CPC Class Codes: H04M 3/4878 Advertisement messages

Abstract

A method of updating or expanding an audio classified ad database with text ads obtained from the databases of print media organizations such as newspapers is disclosed. Ads from a print media ad database are downloaded to a host processor that identifies relevant ads, such as those dealing with autos, and converts the abbreviated text of the print ad to an expanded version of the ad. The words and phrases of the expanded version are parsed and stored in appropriate fields of a relational database. A sequenced play list of audio file numbers corresponding to the words and phrases contained in a database record is created along with "glue" words and phrases that are added to produce a more natural sounding audio. The sequenced list is stored in an audio ad database and is retrieved and assembled to play the audio version of the ad in response to a command from a remote user.

Claims

1. A method for automatically generating an audio ad from a text record, the method comprising the following steps:

receiving an electronic version of the record comprising words and abbreviations of words;

processing the record to expand certain abbreviations of words in the record;

parsing the words and the expanded words from the record based on predetermined lists to create a plurality of audio files;

creating a sequenced list of addresses of audio files corresponding to the parsed words based on a predetermined set of rules; and

generating the audio ad based on the sequenced list of addresses.

2. The method defined in claim 1 comprising the further steps of:

storing the sequenced list of audio file addresses in an audio ad database;

searching the audio ad database for ads matching a request from a remote user; and

composing at least one audio ad using the sequenced list of audio file addresses corresponding to the ads matched in said database.

3. The method defined in claim 1 wherein the record comprises a header and further comprises the step of examining the header of the record to determine whether the record requires processing.

4. The method defined in claim 1 wherein the record comprises a header and further comprises the step of determining a make of an automobile from the header information.

5. The method defined in claim 1 comprising the further step of:

generating an error message when a record is not processed correctly.

6. The method defined in claim 1 wherein creating the sequenced list of addresses further includes inserting a list of addresses of audio files corresponding to glue words between certain words and expanded words to form natural sounding audio ads.

7. A method of generating an audio ad from an electronic ad, the method comprising the steps of:

receiving an electronic ad comprising words;

expanding certain abbreviated words contained in the electronic ad;

inserting glue words before, after, or between one or more words in the expanded electrode ad to generate a modified ad;

identifying known words in the modified ad;

creating a sequenced list of addresses of audio files corresponding to the identified words in the modified ad based on a predetermined set of rules; and

generating the audio ad based on the sequenced list of addresses.

8. The method of claim 7, further comprising the steps of:

identifying search words in the expanded ad, and

storing the search words in a database.

9. The method of claim 7 further comprising the step of:

storing the modified ads in a computer memory to obtain a searchable database of audio ads.

10. The method of claim 9 further comprising the step of:

receiving a search request from a remote user;

searching the database for ads matching the request; and

composing at least one audio ad using the sequenced list of addresses corresponding to the ads matched in the database.