Skip to main content

Boosting Rule-Based Grapheme-to-Phoneme Conversion with Morphological Segmentation and Syllabification in Bengali

  • Conference paper
  • First Online:
Speech and Computer (SPECOM 2023)

Part of the book series: Lecture Notes in Computer Science ((LNAI,volume 14338))

Included in the following conference series:

  • 539 Accesses

Abstract

This paper presents a novel approach to enhance rule-based Bengali grapheme-to-phoneme (G2P) conversion by leveraging morphological segmentation and syllabification techniques. In this approach, input words are first morphologically segmented into valid morphological chunks, each having a different stem and semantic. Applying the G2P rules on each of these chunks, their pronunciations are generated. An intermediate pronunciation for the whole input word is attained by merging these pronunciations. Using syllabification, this intermediate pronunciation is further divided into valid syllabic sequences that offer accurate morphological boundaries. Finally, the final pronunciation is achieved using syllable-specific orthographic rules on these syllabic sequences. The performance of the proposed G2P approach is assessed using measures representing (i) direct accuracy and (ii) enhancements in different speech-related applications. According to the performances noted for the direct accuracy-based measures, the proposed approach predicted the appropriate pronunciations for about 90% cases, especially for compound and inflected words. This performance is around 22% and 10% better than the performance of a rule-based system and a previous state-of-the-art system, respectively. On the other hand, application-based measures guarantee that the generated phone sequences (i) sound natural and (ii) improve the quality of speech synthesis and recognition systems. These quantitative and qualitative assessment plans answered research questions pertinent to speech and linguistics.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Subscribe and save

Springer+ Basic
EUR 32.99 /Month
  • Get 10 units per month
  • Download Article/Chapter or eBook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Chapter
USD 29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
USD 79.99
Price excludes VAT (USA)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
USD 99.99
Price excludes VAT (USA)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Similar content being viewed by others

References

  1. Basu, J., Basu, T., Mitra, M., Mandal, S.K.D.: Grapheme to phoneme (g2p) conversion for bangla. In: 2009 Oriental COCOSDA International Conference on Speech Database and Assessments, pp. 66–71. IEEE (2009)

    Google Scholar 

  2. Bellegarda, J.R.: Unsupervised, language-independent grapheme-to-phoneme conversion by latent analogy. Speech Commun. 46(2), 140–152 (2005)

    Article  Google Scholar 

  3. Bisani, M., Ney, H.: Joint-sequence models for grapheme-to-phoneme conversion. Speech Commun. 50(5), 434–451 (2008)

    Article  Google Scholar 

  4. Black, A.W., Lenzo, K.A.: Building synthetic voices. Lang. Technol. Ins. Carnegie Mellon Univ. Cepstral LLC 4(2), 62 (2003)

    Google Scholar 

  5. Choudhury, M.: Rule-based grapheme to phoneme mapping for Hindi speech synthesis. In: 90th Indian Science Congress of the International Speech Communication Association (ISCA), Bangalore, India. Citeseer (2003)

    Google Scholar 

  6. Dietterich, T.G., Hild, H., Bakiri, G.: A comparative study of id3 and backpropagation for English text-to-speech mapping. In: Machine Learning Proceedings 1990, pp. 24–31. Elsevier (1990)

    Google Scholar 

  7. Ghosh, K., Rao, K.S.: Memory-based data-driven approach for grapheme-to-phoneme conversion in Bengali text-to-speech synthesis system. In: 2011 Annual IEEE India Conference, pp. 1–4. IEEE (2011)

    Google Scholar 

  8. Ghosh, K., Rao, K.S.: Subword based approach for grapheme-to-phoneme conversion in Bengali text-to-speech synthesis system. In: 2012 National Conference on Communications (NCC), pp. 1–5. IEEE (2012)

    Google Scholar 

  9. Ghosh, K., Reddy, R.V., Narendra, N., Maity, S., Koolagudi, S., Rao, K.: Grapheme to phoneme conversion in Bengali for festival based TTS framework. In: 8th International Conference on Natural Language Processing (ICON). Macmillan Publishers (2010)

    Google Scholar 

  10. Ghosh, K., Sreenivasa Rao, K.: Data-driven phrase break prediction for Bengali text-to-speech system. In: Parashar, M., Kaushik, D., Rana, O.F., Samtaney, R., Yang, Y., Zomaya, A. (eds.) IC3 2012. CCIS, vol. 306, pp. 118–129. Springer, Heidelberg (2012). https://doi.org/10.1007/978-3-642-32129-0_17

    Chapter  Google Scholar 

  11. Jiampojamarn, S., Kondrak, G., Sherif, T.: Applying many-to-many alignments and hidden markov models to letter-to-phoneme conversion. In: Human Language Technologies 2007: The Conference of the North American Chapter of the Association for Computational Linguistics; Proceedings of the Main Conference, pp. 372–379 (2007)

    Google Scholar 

  12. Lee, K.F., Hon, H.W., Reddy, R.: An overview of the sphinx speech recognition system. IEEE Trans. Acoust. Speech Signal Process. 38(1), 35–45 (1990)

    Article  Google Scholar 

  13. Lehnert, W.G.: Case-based problem solving with a large knowledge base of learned cases. In: Proceedings of the Sixth National Conference on Artificial Intelligence, vol. 1, pp. 301–306 (1987)

    Google Scholar 

  14. Lucassen, J., Mercer, R.: An information theoretic approach to the automatic determination of phonemic baseforms. In: ICASSP 1984. IEEE International Conference on Acoustics, Speech, and Signal Processing, vol. 9, pp. 304–307. IEEE (1984)

    Google Scholar 

  15. Macherey, K., Dai, A.M., Talbot, D., Popat, A.C., Och, F.: Language-independent compound splitting with morphological operations. In: Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies, HLT 2011, vol. 1, pp. 1395–1404. Association for Computational Linguistics (2011)

    Google Scholar 

  16. Morris, A.C., Maier, V., Green, P.: From WER and RIL to MER and WIL: improved evaluation measures for connected speech recognition. In: Eighth International Conference on Spoken Language Processing, pp. 2765–2768 (2004)

    Google Scholar 

  17. Murthy, H.A., Bellur, A., Viswanath, V., Narayanan, B., Susan, A., Kasthuri, G.: Building unit selection speech synthesis in Indian languages: an initiative by an Indian consortium. In: Proceedings of COCOSDA (2010)

    Google Scholar 

  18. Narendra, N., Rao, K.S., Ghosh, K., Reddy, V.R., Maity, S.: Development of bengali screen reader using festival speech synthesizer. In: 2011 Annual IEEE India Conference, pp. 1–4. IEEE (2011)

    Google Scholar 

  19. Narendra, N., Rao, K.S., Ghosh, K., Vempada, R.R., Maity, S.: Development of syllable-based text to speech synthesis system in Bengali. Int. J. Speech Technol. 14, 167–181 (2011)

    Article  Google Scholar 

  20. Oakey, S., Cawthorn, R.: Inductive learning of pronunciation rules by hypothesis testing and correction. In: IJCAI, pp. 109–114. Citeseer (1981)

    Google Scholar 

  21. Reichel, U.D., Schiel, F.: Using morphology and phoneme history to improve grapheme-to-phoneme conversion. In: Proceedings of the Eurospeech, pp. 1937–1940 (2005)

    Google Scholar 

  22. Stanfill, C.: Memory-Based Reasoning Applied to English Pronunciation. Thinking Machines Corporation, Cambridge (1987)

    Google Scholar 

  23. Van Coile, B.: Inductive learning of pronunciation rules with the depes system. In: IEEE International Conference on Acoustics, Speech, and Signal Processing, pp. 745–748. IEEE Computer Society (1991)

    Google Scholar 

  24. Wagner, R.A., Fischer, M.J.: The string-to-string correction problem. J. ACM (JACM) 21(1), 168–173 (1974)

    Article  MathSciNet  MATH  Google Scholar 

Download references

Acknowledgement

We would like to express our sincere gratitude to Mr Prabhat Mukherjee for his contribution in collecting data and generating the pronunciation dictionary and Mr Nilay Roy for his guidance as a linguist. We also acknowledge the collaborative spirit of the students of KIIT University in the evaluation tasks.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Krishnendu Ghosh .

Editor information

Editors and Affiliations

Rights and permissions

Reprints and permissions

Copyright information

© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG

About this paper

Check for updates. Verify currency and authenticity via CrossMark

Cite this paper

Ghosh, K., Mandal, S., Roy, N. (2023). Boosting Rule-Based Grapheme-to-Phoneme Conversion with Morphological Segmentation and Syllabification in Bengali. In: Karpov, A., Samudravijaya, K., Deepak, K.T., Hegde, R.M., Agrawal, S.S., Prasanna, S.R.M. (eds) Speech and Computer. SPECOM 2023. Lecture Notes in Computer Science(), vol 14338. Springer, Cham. https://doi.org/10.1007/978-3-031-48309-7_34

Download citation

  • DOI: https://doi.org/10.1007/978-3-031-48309-7_34

  • Published:

  • Publisher Name: Springer, Cham

  • Print ISBN: 978-3-031-48308-0

  • Online ISBN: 978-3-031-48309-7

  • eBook Packages: Computer ScienceComputer Science (R0)

Publish with us

Policies and ethics