Exploring the impact of error type features integration on transformer based Myanmar spelling correction

Name: Ei, Thandar Phyu

Organization : Kasetsart University. Faculty of Engineering

Email : eithandar.p@ku.th

Name: Ye, Kyaw Thu

Organization : National Electronics and Computer Technology Center

Email : yekyaw.thu@nectec.or.th

Name: Hutchatai Chanlekha

Organization : Kasetsart University. Faculty of Engineering

Email : fenghtc@ku.ac.th

Name: Funakoshi, Kotaro

Organization : Tokyo Institute of Technology

Email : funakoshi@lr.pi.titech.ac.jp

Name: Thepchai Supnithi

Organization : National Electronics and Computer Technology Center

Email : thepchai.supnithi@nectec.or.th

keyword: Spelling correction

LCSH: Burmese language -- Orthography and spelling

; Transformers (Machine learning)

LCSH: Natural language processing (Computer science)

LCSH: Error analysis (Mathematics)

LCSH: Computational linguistics

Abstract: Spelling correction is a fundamental aspect of natural language processing systems, essential for ensuring the accuracy and fluency of textbased applications. In the context of neural machine translation based spelling correction, spelling errors pose significant challenges, particularly in lowresource language like Myanmar language where training data is limited. In this paper, the Myanmar spelling errorcorrection parallel corpus was developed through the collection of text from social media platforms. By learning the error patterns in social media text data, we categorized the Myanmar spelling mistakes into ten categories. We developed Myanmar spelling correction model using Transformer architecture and also investigated the potential improvement of our Transformerbased spelling correction by incorporating error type features as additional linguistic information. Evaluating the Transformer model with manually annotated error type labels in our developed corpus significantly enhances the F1score from 75.91% to 86.71%, increases the GLEU score from 0.9648 to 0.9767, and decreases the word error rate from 3.3% to 2.0%. Additionally, we conducted a preliminary study on integrating our proposed correction model with an automatic error type classification model, employing a bidirectional Long ShortTerm Memory. The findings demonstrate that integrating error type information into the Transformer architecture enhances the effectiveness of Myanmar spelling correction.

King Mongkut's University of Technology North Bangkok. Central Library

Address: BANGKOK

Email: library@kmutnb.ac.th

Created: 2024

Modified: 2025-05-27

Issued: 2025-05-27

บทความ/Article

application/pdf

BibliograpyCitation : In IEEE Thailand Section (IEEE Computer Society Thailand Chapter) and Prince of Songkla University. College of Computing. The 21st International Joint Conference on Computer Science and Software Engineering (JCSSE 2024)) (pp.365-372). Phuket : Prince of Songkla University

eng

©copyrights King Mongkut's University of Technology North Bangkok

ลำดับที่.	ชื่อแฟ้มข้อมูล	ขนาดแฟ้มข้อมูล	จำนวนเข้าถึง	วัน-เวลาเข้าถึงล่าสุด
1	JCSSE 2024pp.365-372.pdf	248.73 KB

ใช้เวลา

0.017892 วินาที