Voice dialog server method and system

• US 6,999,930 B1
• Filed: 03/27/2002
• Issued: 02/14/2006
• Est. Priority Date: 03/27/2002
• Status: Active Grant

• Current Assignee: Extended Systems Inc. (SAP SE)
• Original Assignee: Extended Systems Inc. (SAP SE)
• Inventors: Holsinger, David, Roberts, Steve
• Primary Examiner(s): Dorvil, Richemond
• Assistant Examiner(s): Han, Qi
• Application Number:
  US10/113,710
• Time in Patent Office: 1,420 Days
• Field of Search: 704/270.1, 704/257, 704/1, 704/231, 704/235, 704/9, 379/88.01, 379/88.1
• US Class Current: 704/270.1
• CPC Class Codes: G10L 15/193 Formal grammars, e.g. finite state automata, context free grammars or word networks
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
• CPC Class Codes: H04M 2201/39 using speech synthesis
• CPC Class Codes: H04M 2201/40 using speech recognition
• CPC Class Codes: H04M 3/4938 comprising a voice browser which renders and interprets, e.g. VoiceXML

Abstract

A voice user interface for accessing data is provided. This method and system receives a set of variables from a voice recognizer processing a voice request for information in a backend system; processes an integration method using the set of variables returned from the recognizer and an unstructured grammar based on entries in the database fields of a backend system; requests backend data from the backend system using the integration method and the set of variables as applied to the unstructured grammar; and receives backend data from the backend system responsive to the integration method used to request the backend data. Processes are described that add data associated with a database to increase the accuracy of interacting with the database through a voice interface and create the unstructured grammar for improved voice recognition capabilities.

Claims

1. A method for accessing data from entries in a database fields of a backend system, comprising:

providing an unstructured grammar generated, at least in part, from the entries in the database fields of the backend system, the unstructured grammar including permutations of data in the various entries in those database fields;

updating the unstructured grammar as the database of the backend system is modified;

receiving a set of variables from a voice recognizer processing a voice request for information in the backend system;

processing an integration method using the set of variables returned from the recognizer and the unstructured grammar generated from entries in the database fields of a backend system;

requesting backend data from the backend system using the integration method and the set of variables as applied to the unstructured grammar; and

receiving backend data from the backend system responsive to the integration method used to request the backend data.

2. The method of claim 1 further comprising, ranking the received data using extra-grammatical information to increase the probability that the backend data provided is responsive to the voice request.

3. The method of claim 2 wherein the extra-grammatical information includes calendar information to increase the probability that the backend data provided is responsive to the voice request.

4. The method of claim 2 wherein the extra-grammatical information includes location-based information to increase the probability that the backend data provided is responsive to the voice request.

5. The method of claim 2 wherein the extra-grammatical information includes statistical information on a user or group of users to increase the probability that the backend data provided is responsive to the voice request.

6. The method of claim 2 wherein the extra-grammatical information includes security and authorization information associated with a user to increase the probability that the backend data provided is responsive to the voice request.

7. The method of claim 2 wherein the extra-grammatical information used to rank the data from the backend system includes a set of user preferences.

8. The method of claim 2 wherein the extra-grammatical information used to rank the data from the backend system includes historical voice request information by a user.

9. The method of claim 2 wherein the extra-grammatical information used to rank the data from the backend system includes historical categorical information requested by a user.

10. The method of claim 2 wherein the extra-grammatical information used to rank the data from the backend system includes a list of entities having previous relationships with the user.

11. The method of claim 1 further comprising, processing a state model to assist in processing the integration method corresponding to the voice request for data from the backend system.

12. The method of claim 11 wherein the state model further describes the manner of presenting data from the backend system.

13. The method of claim 1 wherein the integration method is compatible with a database query language used to access data in a database.

14. The method of claim 1 wherein the unstructured grammar corresponds to a permutation of entries in one or more fields in a database associated with a backend system.

15. The method of claim 14 wherein the entries in one or more fields in the database further include additional identifiers that correspond and reference one or more entries in the database.

16. The method of claim 15 wherein the additional identifiers are selected from a set of identifier types including abbreviations, alternate abbreviations, synonyms, homonyms, misspellings, slang, vernacular and other identifiers associated with each of the one or more entries in one or more fields in the database.

17. The method of claim 1, wherein providing includes generating the unstructured grammar wherein generating comprises:

combining data entries from data fields in the database of the backend system; and

generating permutations of the data entries to increase the overall size of the grammar and probability of identifying matches against the grammar from the voice requests.

18. The method of claim 17, further comprising prioritizing the permutations of the data entries according to information in the database, wherein the permutations of the data entries are prioritized according to extra-grammatical information related to the voice request.

19. The method of claim 17, further comprising prioritizing the permutations of the data entries according to information in the database, wherein the permutations of the data entries are prioritized according to extra-grammatical information related to a user.

20. The method of claim 17, further comprising prioritizing the permutations of the data entries according to information in the database, wherein the permutations of the data entries are prioritized according to extra-grammatical information describing historical patterns of accessing data entries in the database.

21. The method of claim 17 wherein the database is associated with a backend system used to manage enterprise information.

22. The method of claim 17 wherein the data entries used to create the unstructured grammar are selected from a subset of data fields in the database associated with extra-grammatical information.