Augmented words to improve a deep learning-based Indonesian syllabification

Abstract

Recent deep learning-based syllabification models generally give low error rates for high-resource languages with big datasets but sometimes produce high error rates for the low-resource ones. In this paper, two procedures: massive data augmentation and validation, are proposed to improve a deep learning-based syllabification, using a combination of bidirectional long short-term memory (BiLSTM), convolutional neural networks (CNN), and conditional random fields (CRF) for a low-resource Indonesian language. The massive data augmentation comprises four methods: transposing nuclei, swapping consonant-graphemes, flipping onsets, and creating acronyms. Meanwhile, the validation is implemented using a phonotactic-based scheme. A preliminary investigation on 50k Indonesian words informs that those augmentation methods significantly enlarge the dataset size by 12.8M valid words based on the phonotactic rules. An examination is then performed using 5-fold crossvalidation. It reports that the augmentation methods significantly improve the BiLSTM-CNN-CRF model for 50k formal words and 100k named-entities datasets. A detailed investigation informs that augmenting the training set can reduce the word error rate (WER) coming from the long formal words and named entities.

Evaluating the Sesotho rule-based syllabification system on Sepedi and Setswana words

“…According to Suyanto et al (2021), rule-based systems perform better for low-resourced languages with limited to no gold standard corpora. This was also demonstrated in the case of Sesotho where the rule-based system outperformed the T E Xbased patterning system (Sibeko and van Zaanen, 2022a).…”