Understanding without formality: augmenting speech recognition to understand informal verbal commands

ABSTRACT

Verbal command and control systems are fairly common; almost all off-the-shelf speech recognition packages come with a way to perform various tasks through a voice command. Unfortunately, these systems require that the user utter the commands precisely in the format that it is expecting. These systems have a small number of grammar rules defined that are used to match against incoming utterances. Here, we present a method of using these same grammar rules to expand the capabilities of command and control engines to include semantically similar utterances. Latent Semantic Analysis (LSA) is used to connect specific grammar rules with the meanings underlying matching phrases resulting in utterances being matched to grammar rules even though the exact phrase did not match any specific rule. Experiments are described that determine the extent to which this method can be used and how accurate it is.

REFERENCES

• 1: CNN online news, http://www.cnn.com, 2004.
• 2: Resource.com, http://thesaurus.reference.com, 2004.
• 3: L. Douglas Baker , Andrew Kachites McCallum, Distributional clustering of words for text classification, Proceedings of the 21st annual international ACM SIGIR conference on Research and development in information retrieval, p.96-103, August 24-28, 1998, Melbourne, Australia [doi>10.1145/290941.290970]
• 4: Bellegarda, J. R., Butzberger, J. W., Chow, Y. L., Coccaro, N. and Naik, D., Automatic Discovery of Word Classes Through Latent Semantic Analysis, in EUSIPCO-96 Signal Processing VIII, Theories and Applications, (1996), Edizioni Lint Trieste.
• 5: Bellegarda, J. R., Butzberger, J. W., Chow, Y. L., Coccaro, N. and Naik, D., A Novel Word Clustering Algorithm Based on Latent Semantic Analysis, in ICASSP-96, (1996), 172--175.
• 6: Burgess, C., Livesay, K. and Lund, K. Explorations in Context Space: Words, Sentences, Discourse. Discourse Processes, 25. 211--257.
• 7: Deerwester, S., Dumais, S. T., Fumas, G. W., Landaur, T. K. and Harshman, R. Indexing by latent semantic analysis. Journal of the American Society for Information Science, 41. 391--407.
• 8: Dumais, S. T. Latent semantic indexing (LSI) and TREC-2. in Harman, D. ed. National Institute of Standards and Technology Text Retrieval Conference, NIST, 1994.
• 9: Yoshihiko Gotoh , Steve Renals, Topic-based mixture language modelling, Natural Language Engineering, v.5 n.4, p.355-375, December 1999 [doi>10.1017/S1351324900002278]
• 10: Khudanpur, S. and Wu, J., A maximum entropy language model integrating n-grams and topic dependencies for conversational speech recognition. in ICASSP-99, (Phoenix, AZ, 1999), 553--556.
• 11: R. Kuhn , R. De Mori, A Cache-Based Natural Language Model for Speech Recognition, IEEE Transactions on Pattern Analysis and Machine Intelligence, v.12 n.6, p.570-583, June 1990 [doi>10.1109/34.56193]
• 12: Landaur, T. K. and Dumais, S. T. A solution to Plato's problem: The Latent Semantic Analysis theory of the acquisition, induction, and representation of knowledge. Psychological Review, 104. 211--240.
• 13: Rea, L. Designing and Conducting Survey Research. Jossey-Bass, 1997.
• 14: Siivola, V., Language modeling based on neural clustering of words, in IDIAP-Com 02, (Martigny, Switzerland, 2000).

INDEX TERMS

Primary Classification:
  H. Information Systems
  H.5 INFORMATION INTERFACES AND PRESENTATION (I.7)
      H.5.2 User Interfaces (D.2.2, H.1.2, I.3.6)

Additional Classification:
  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE
      I.2.7 Natural Language Processing

General Terms:
Experimentation, Human Factors, Reliability

Keywords:
command-and-control, latent semantic analysis, natural language understanding, speech recognition