Pantun Bermukun Sarawak: Automated Pantun Reply Generation using Fuzzy N-gram Similarity

Pantun bermukun, an interactive oral poetry tradition of the Sarawak Malay community, is under threat due to declining practice and the lack of dialogic paired pantun data. Existing collections predominantly consist of single-verse pantun, impeding both digital preservation and computational analysi...

Full description

Saved in:
Bibliographic Details
Main Authors: Mohammad, Hossin, Hamizan, Sharbini, Fatimah, Hj Subet
Format: Proceeding
Language:en
Published: 2025
Subjects:
Online Access:http://ir.unimas.my/id/eprint/51388/1/THE%2010th%20INTERNATIONAL.pdf
http://ir.unimas.my/id/eprint/51388/
https://cdnc.heyzine.com/files/uploaded/2c6d8b7eacb4917421f98f25e258b2e60a9523fc.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Pantun bermukun, an interactive oral poetry tradition of the Sarawak Malay community, is under threat due to declining practice and the lack of dialogic paired pantun data. Existing collections predominantly consist of single-verse pantun, impeding both digital preservation and computational analysis. This paper proposes an automated reply selection method that utilises N-gram similarity, augmented with fuzzy matching and Levenshtein distance, to identify the most appropriate replypantun from a dataset of 316 single-verse Pantun Melayu Sarawak (PMS). Evaluation by five subject matter experts (SMEs) on 50 query pantun indicates that the fuzzy matching-enhanced approach outperforms the standard Jaccard coefficient method in generating suitable replies (80.4%vs. 74.8%). The results demonstrate the viability of automated pantun pairing for digital preservation, while also highlighting the limitations of current text similarity measures in handling semantic and culturally nuanced pantun. These findings support the development of computational tools for the documentation and revitalisation of pantun bermukun traditions.