| US 7,467,090 B1 | ||
| Using web FAQ data for creating self-service speech applications | ||
| Osamuyimen T. Stewart, Piscataway, N.J. (US); David M. Lubensky, Brookfield, Conn. (US); Ea-Ee Jan, Ardsley, N.Y. (US); and Xiang Li, Milpitas, Calif. (US) | ||
| Assigned to International Business Machines Corporation, Armonk, N.Y. (US) | ||
| Filed on Feb. 27, 2008, as Appl. No. 12/37,996. | ||
| Int. Cl. G10L 21/00 (2006.01) | ||
| U.S. Cl. 704—270 [704/270.1; 704/275; 379/88.01] | 1 Claim |
| 1. A method of using web data for telephone self-service, via a speech application, said method comprising:
pulling scripts for frequently-asked issues for all problem types from a relevant web domain using a web crawler;
automatically parsing a content of said frequently-asked issues into speech acts relevant to a speech domain, by picking out
verbs, verb phrases, nouns, noun phrases, adjective phrases, or adjunct clauses;
reformatting said speech acts;
creating a different number of steps corresponding to said speech acts,
wherein a causality function is preserved, by keeping a sequential order of said verbs associated with said speech acts;
analyzing all said speech acts corresponding to a specific problem;
categorizing all said speech acts corresponding to said specific problem into multiples numbers of classes by counting, corresponding
to different ranges of complexity of steps, based on a typical human's short-term memory limitations and recollection, wherein
the classes comprise simple, medium, complex, and complex-complex;
playing back a script of a content corresponding to said specific problem, according to said categorized classes, problem
type of said specific problem, and frequently-asked issue associated with said specific problem; and
providing a user interface for a user, to enable said user control and navigate said play back.
|