Aqsa, Aqila (2020) Pengembangan Sistem Interaktif & Inklusif Berbasis Website Sebagai Pendukung Pembelajaran Sosial-emosional Anak Usia Dasar Dan Anak Dengan Autisme Menggunakan Natural Language Processing Dan Natural Language Generation. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5027211032-Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5027211032-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (8MB) | Request a copy |
Abstract
Kesulitan dalam mengenali dan mengekspresikan emosi merupakan tantangan umum bagi anak usia sekolah dasar, khususnya anak dalam spektrum autisme. Kompetensi sosial emosional yang mendukung interaksi sosial dan regulasi diri dapat ditingkatkan melalui pendekatan edukatif berbasis cerita yang relevan dengan kehidupan sehari-hari. Penelitian ini mengembangkan sistem web interaktif berbasis narasi visual yang dirancang untuk membantu anak mengenali enam emosi dasar melalui teknologi Natural Language Processing (NLP) dan Natural Language Generation (NLG). Sistem ini menggunakan model IndoBERT untuk klasifikasi emosi dan GPT-4 untuk menghasilkan umpan balik edukatif secara adaptif. Dataset dikurasi dari dua sumber Bahasa Inggris yang diterjemahkan ke Bahasa Indonesia dan digunakan untuk melatih model NLP yang menunjukkan kinerja unggul dengan F1-score tertinggi sebesar 91,4% dan akurasi 90,8%, melebihi capaian metrik pada acuan penelitian terdahulu oleh Patil et al. dan Alatrash et al. yang berkisar di bawah 90% untuk kedua metrik. Sistem ini juga dilengkapi fitur speech-to-text, elemen gamifikasi, serta antarmuka ramah anak yang mendukung keterlibatan pengguna. Hasil pengujian menunjukkan bahwa sistem ini inklusif, mudah digunakan, dan efektif dalam meningkatkan kesadaran emosional dan regulasi diri anak, serta memberikan pengalaman belajar yang menyenangkan dan adaptif.
==================================================================================================================================
Difficulties in recognizing and expressing emotions are common challenges among elementary school-aged children, particularly those on the autism spectrum. Social-emotional competence, which supports social interaction and self regulation, can be enhanced through educational approaches based on narrative scenarios relevant to everyday life. This study developed an interactive web-based system using visual narratives designed to help children identify six basic emotions through Natural Language Processing (NLP) and Natural Language Generation (NLG) technologies. The system employs the IndoBERT model for emotion classification and GPT-4 to generate adaptive educational feedback. The dataset was curated from two English-language sources and translated into Indonesian to train the NLP model, which demonstrated strong performance with a highest F1-score of 91.4% and accuracy of 90.8%—exceeding the metric outcomes reported in prior studies by Patil et al. and Alatrash et al., which remained below 90% for both measures. The system also integrates speech-to-text functionality, gamified elements, and a child-friendly interface to support user engagement. Testing results indicate that the system is inclusive, easy to use, and effective in enhancing children's emotional awareness and self-regulation, while offering an engaging and adaptive learning experience.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Pembelajaran Sosial-Emosional, Natural Language Processing (NLP), Natural Language Generation (NLG), GPT, IndoBERT, Social-Emotional Learning, Natural Language Processing (NLP), Natural Language Generation (NLG), GPT, IndoBERT |
| Subjects: | T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T58.6 Management information systems T Technology > T Technology (General) > T58.8 Productivity. Efficiency |
| Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information Technology > 59201-(S1) Undergraduate Thesis |
| Depositing User: | Aqila Aqsa |
| Date Deposited: | 22 Jul 2025 02:08 |
| Last Modified: | 22 Jul 2025 02:08 |
| URI: | http://repository.its.ac.id/id/eprint/120411 |
Actions (login required)
 |
View Item |