Knowledge-based strategies applied to N-best lists in automatic speech recognition systems
A highly accurate technique for recognizing spoken digit strings is described. A spoken digit string is received and analyzed by a speech recognizer, which generates a list of hypothesized digit strings arranged in ranked order based on a likelihood of matching the spoken digit string. The individual hypothesized strings are then analyzed in order beginning with the hypothesized string having the greatest likelihood of matching the spoken string to determine whether they satisfy a given constraint. The first hypothesized string in the list satisfying the constraint is selected as the recognized string.
1. A method of recognizing a spoken digit string, comprising:
(a) receiving the spoken digit string;
(b) analyzing the spoken digit string to generate a list of hypothesized digit strings arranged in ranked order based on a likelihood of matching the spoken digit string;
(c) determining whether individual hypothesized strings of aid list satisfy a given constraint, using a given knowledge based recognition strategy;
(d) selecting the first string in the list satisfying the constraint as the recognized string, if none of the hypothesized digit strings satisfy the constraint,
(e) prompting entry of a repeated spoken digit string, which is a an entire repeat of the spoken digit string entered in step (a);
(f) analyzing the repeated spoken digit string to generate a second list of hypothesized digit strings arranged in ranked order based on a likelihood of matching the repeated spoken digit string;
(g) selecting the recognized string in accordance with a comparison of the first and second list if the constraint is satisfied;
(h) using the first list as a first list database to verify one of the hypothesized strings in the second list;
(i) selecting the recognized string in the second list that was verified by the first list; or
(j) if there is no recognized string in the second list verified by the first list, then performing additional verification techniques to determine the correct digit string until the constraint is satisfied, and then subsequently selecting the correct digit string, wherein said additional verification techniques include performing a least one of a checksum approach with the N-best list and a personal identification number, a database match comparison with valid entries, digit positional constraints and fuzzy matching criterion.
2. The method of claim 1 wherein said given knowledge based recognition strategy comprises a database matching scheme.
3. The method of claim 2 , wherein step (c) comprises searching a valid data strings database to determine whether any of the hypothesized digit strings match one of the valid digit strings.
4. The method of claim 1 wherein the knowledge based recognition strategy is a checksum scheme.
5. The method of claim 4 wherein the spoken digit string includes a checksum digit, and wherein step (c) comprises calculating a checksum of the hypothesized digit strings and determining whether the checksum matches the value of the checksum digit.
6. The method of claim 4 wherein the checksum scheme utilized a Luhn checksum algorithm.
7. The method of claim 1 wherein the knowledge based recognition strategy is a digit positional strategy and the constraining is a given digit position.
8. The method of claim 1 wherein the knowledge based recognition strategy is a digit string length strategy and the constraint is a given digit string length.
9. The method of claim 1 further comprising the step of prompting entry of a spoken digit string prior to its receipt in step (a).
10. The method of claim 1 , wherein step (h) further comprises
(ii) if there is no recognized string verified in step (i), then using the second list as a database to verify one of the hypothesized strings in the first list; and
(iii) selecting the recognized string in the second list that was verified by the first list.