Claim Missing Document
Check
Articles

Found 2 Documents
Search

Fine-Tuning Whisper Model for Mandar Speech Recognition: Approach and Performance Evaluation Jafar Jafar; Mar Athul Wazithah Tb; Firman Aziz; Rosary Iriany; Norma Nasir
Journal of Applied Engineering and Technological Science (JAETS) Vol. 7 No. 1 (2025): Journal of Applied Engineering and Technological Science (JAETS)
Publisher : Yayasan Riset dan Pengembangan Intelektual (YRPI)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37385/jaets.v7i1.7170

Abstract

This research focuses on the development of speech recognition technology for the Mandar language, a regional language in Indonesia with limited digital resources. The main challenge lies in the lack of local datasets and the minimal representation of the Mandar language in existing multilingual speech recognition models. This study aims to enhance the performance of Automatic Speech Recognition (ASR) systems by fine-tuning the Whisper model using a Mandar-specific dataset. The dataset consists of 1,000 audio recordings with various dialects and recording qualities, which underwent preprocessing steps such as segmentation, normalization, and data augmentation. Fine-tuning was conducted using supervised learning methods with hyperparameter optimization, resulting in a reduction of Word Error Rate (WER) from 73.7% in the pretrained model to 37.4% after fine-tuning, and an increase in accuracy from 26.3% to 62.6%. The optimized model was also compared with other ASR models, such as DeepSpeech and wav2vec 2.0, demonstrating superior performance in terms of accuracy and time efficiency. Further analysis revealed that recording quality and dialect variations significantly impacted model performance, with high-quality recordings and standard dialects yielding the best results. The model was implemented as a web application prototype, enabling efficient and near real-time transcription of Mandar speech. This research not only contributes to the development of ASR technology for low-resource languages but also opens new opportunities for preserving and utilizing the Mandar language through digital technology. For future improvements, larger datasets, more advanced augmentation techniques, and the exploration of additional language model integration are recommended.
Pelatihan Penyusunan Asesmen Literasi Numerasi Terintegrasi dalam Pembelajaran Tematik Bagi Guru SD di UPT SD Negeri 20 Bangkala Barat Jafar Jafar; Mantasiah R; Rizki Herdiani; Norma Nasir; Fathullah Wajdi
Jurnal Pengabdian Kepada Masyarakat Patikala Vol. 6 No. 1 (2026): Jurnal PkM PATIKALA
Publisher : Pusat Pengembangan Pendidikan dan Bakat Indonesia/Education and Talent Development Center of Indonesia (ETDC Indonesia)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51574/patikala.v6i1.5669

Abstract

Numeracy literacy is an essential competency that needs to be developed starting at the primary school level through instruction and assessment capable of measuring reasoning skills and the ability to solve contextual problems. However, teachers still face challenges in designing numeracy literacy assessments that are integrated with thematic learning. This community service activity aimed to enhance teachers' knowledge and skills in developing numeracy literacy assessments aligned with the characteristics of thematic learning. The activity was conducted at UPT SD Negeri 20 Bangkala Barat, Jeneponto Regency, involving 11 teachers as participants. A participatory training method was employed, comprising four stages: preparation, training implementation, workshops and mentoring, and evaluation and reflection. Evaluation was carried out using pre-tests and post-tests, assessments of the assessment products created by participants, and participant response questionnaires. The results showed an increase in the participants' average score from 61.82 in the pre-test to 87.36 in the post-test. The resulting assessment products achieved an average score of 3.60 (categorized as "very good"), while participant response to the activity implementation averaged 96.9% (also categorized as "very good"). In addition to improving conceptual understanding, this activity produced examples of numeracy literacy assessment tools ready for classroom implementation. The program demonstrates that training combined with workshops and mentoring is effective in enhancing teachers' competence in designing numeracy literacy assessments. Similar activities should be conducted sustainably and expanded to more schools to generate a broader impact on improving the quality of learning in primary schools.