20.500.12556/DKUM-91741 Sequence-to-Sequence models and their evaluation for spoken language normalization of Slovenian Sequence-to-sequence models have been applied to many challenging problems, including those in text and speech technologies. Normalization is one of them. It refers to transforming non-standard language forms into their standard counterparts. Non-standard language forms come from different written and spoken sources. This paper deals with one such source, namely speech from the less-resourced highly inflected Slovenian language. The paper explores speech corpora recently collected in public and private environments. We analyze the efficiencies of three sequence-to-sequence models for automatic normalization from literal transcriptions to standard forms. Experiments were performed using words, subwords, and characters as basic units for normalization. In the article, we demonstrate that the superiority of the approach is linked to the choice of the basic modeling unit. Statistical models prefer words, while neural network-based models prefer characters. The experimental results show that the best results are obtained with neural architectures based on characters. Long short-term memory and transformer architectures gave comparable results. We also present a novel analysis tool, which we use for in-depth error analysis of results obtained by character-based models. This analysis showed that systems with similar overall results can differ in the performance for different types of errors. Errors obtained with the transformer architecture are easier to correct in the post-editing process. This is an important insight, as creating speech corpora is a time-consuming and costly process. The analysis tool also incorporates two statistical significance tests: approximate randomization and bootstrap resampling. Both statistical tests confirm the improved results of neural network-based models compared to statistical ones. low-resource language applications spoken language normalization character unit subword unit statistical model long short-term memory transformer error analysis aplikacije govorjeni jeziki statistični modeli true false true MDPI Angleški jezik Slovenski jezik https://www.mdpi.com/journal/applsci © 2024 by the authors Članek v reviji 2025-01-31 11:07:53 2025-01-31 11:07:53 2025-08-27 11:55:41 0000-00-00 00:00:00 2024 0 0 24 str. št. 20, št. članka 9515 let. 14 okt. 2024 0000-00-00 Zaloznikova Objavljeno NiDoloceno 2024-09-06 2024-10-16 2024-10-18 004.8 213048067 10.3390/app14209515 2076-3417 applsci-14-09515.pdf applsci-14-09515.pdf 1 19D67562D60D137292969407CD398506 a16c7fcb0d11658d99abebfda27b4c41779c51d409abfd447ba6ef3f48248468 6946e0df-dfbb-11ef-ae92-00155d000105 https://dk.um.si/Dokument.php?lang=slv&id=188367 https://www.mdpi.com/2076-3417/14/20/9515 0 fd8d1b3b-832b-11f0-8e8c-00155d000105 https://dk.um.si/Dokument.php?lang=slv&id=196578 Fakulteta za elektrotehniko, računalništvo in informatiko 0 0 0