Method and apparatus for generating hint words for automated speech recognition
Systems and methods for determining hint words that improve the accuracy of automated speech recognition (ASR) systems. Hint words are determined in the context of a user issuing voice commands in connection with a voice interface system. Terms are initially taken from most frequently occurring terms in operation of a voice interface system. For example, most frequently occurring terms that arise in electronic search queries or received commands are selected. Certain of these terms are selected as hint words, and the selected hint words are then transmitted to an ASR system to assist in translation of speech to text.
1. A method of determining hint words for automated speech recognition, the method comprising:
determining, using processing circuitry, a first set of terms comprising terms that are most frequently occurring terms from operation of a voice interface system, the most frequently occurring terms selected from one or more of terms of queries issued to the voice interface system, or terms of commands issued to the voice interface system;
determining, using the processing circuitry, a second set of terms that are most frequently occurring terms arising during a predetermined time period of operation of the voice interface system;
selecting as a first set of hint words facilitating operation of an automated speech recognition application, the first set of hint words comprising common terms of the first set of terms and the second set of terms;
selecting, if less than a predetermined number of common terms are selected, a second set of hint words comprising a plurality of terms from the first set of terms that are not among the common terms; and
transmitting, using the processing circuitry, the first set of hint words and the second set of hint words to the automated speech recognition application.
2. The method of claim 1 , wherein the second set of hint words further comprises selecting a predetermined number of most frequently occurring ones of the terms as the hint words.
3. The method of claim 1 , wherein the second set of hint words further comprises a first predetermined number of the first set of terms and a second predetermined number of the second set of terms.
4. The method of claim 1 ,
wherein a sum of the number of first hint words and a number of the second hint words is equal to the predetermined number.
5. The method of claim 1 ,
wherein the second hint words further comprises one or more terms from the second set of terms that are not among the common terms and
wherein a sum of the number of first hint words and a number of the second hint words is equal to the predetermined number.
6. The method of claim 1 , wherein the most frequently occurring terms are selected from one or more of terms of most recent queries issued to the voice interface system, or terms of most recent commands issued to the voice interface system.
7. The method of claim 1 , wherein the most frequently occurring terms are selected from one or more of terms of queries issued to the voice interface system, terms of commands issued to the voice interface system, or phonemes thereof.
8. The method of claim 1 , wherein the most frequently occurring terms are selected from one or more of terms of queries issued to the voice interface system, terms of commands issued to the voice interface system, or phonetic neighbors thereof.
9. The method of claim 1 , wherein at least one of the terms of queries or the terms of commands comprise one or more of names of consumer goods, tasks, reminders, calendar items, dates, or items of a list of items.
10. A system for determining hint words for automated speech recognition, the method comprising:
a storage device; and
processing circuitry configured to:
determine a first set of terms comprising terms that are most frequently occurring terms from operation of a voice interface system, the most frequently occurring terms selected from one or more of terms of queries issued to the voice interface system, or terms of commands issued to the voice interface system;
determine a second set of terms that are most frequently occurring terms arising during a predetermined time period of operation of the voice interface system;
select as a first set of hint words facilitating operation of an automated speech recognition application, the first set of hint words comprising common terms of the first set of terms and the second set of terms; and
select, if less than a predetermined number of common terms are selected, a second set of hint words comprising a plurality of terms from the first set of terms that are not among the common terms; and
input/output circuitry configured to:
transmit the first set of hint words and the second set of hint words to the automated speech recognition application.
11. The system of claim 10 , wherein the selecting further comprises selecting a predetermined number of the most frequently occurring ones of the terms as the hint words.
12. The system of claim 10 , wherein the second set of hint words further comprises a first predetermined number of the first set of terms and a second predetermined number of the second set of terms.
13. The system of claim 10 ,
wherein a sum of the number of first hint words and a number of the second hint words is equal to the predetermined number.
14. The system of claim 10 ,
wherein the second hint words further comprises one or more terms from the second set of terms that are not among the common terms and
wherein a sum of the number of first hint words and a number of the second hint words is equal to the predetermined number.
15. The system of claim 10 , wherein the most frequently occurring terms are selected from one or more of terms of most recent queries issued to the voice interface system, or terms of most recent commands issued to the voice interface system.
16. The system of claim 10 , wherein the most frequently occurring terms are selected from one or more of terms of queries issued to the voice interface system, terms of commands issued to the voice interface system, or phonemes thereof.
17. The system of claim 10 , wherein the most frequently occurring terms are selected from one or more of terms of queries issued to the voice interface system, terms of commands issued to the voice interface system, or phonetic neighbors thereof.
18. The system of claim 10 , wherein at least one of the terms of queries or the terms of commands comprise one or more of names of consumer goods, tasks, reminders, calendar items, dates, or items of a list of items.
19. A method of determining hint words for automated speech recognition, the method comprising:
determining, using processing circuitry, a first set of terms comprising most frequently occurring terms selected from one or more of terms of queries issued to a voice interface system or terms of commands issued to the voice interface system;
determining, using the processing circuitry, a second set of terms that are most frequently occurring terms arising during a predetermined time period of operation of the voice interface system;
selecting as a first set of hint words facilitating operation of an automated speech recognition application, the first set of hint words comprising common terms of the first set of terms and the second set of terms;
selecting a second set of hint words comprising a plurality of terms from that are not among the common terms from at least one of the first set of terms or the second set of terms; and
transmitting, using the processing circuitry, the hint words to the automated speech recognition application, the hint words comprising the first set of hint words and the second set of hint words.