Please use this identifier to cite or link to this item:
http://repository.ipb.ac.id/handle/123456789/165965| Title: | Pengembangan Chatbot Berbasis Retrieval Augmented Generation dengan Studi Kasus Dasar Pokok Agraria |
| Other Titles: | Chatbot Development Based on Retrieval Augmented Generation with Basic Agrarian Case Study |
| Authors: | Hermadi, Irman Sururi, Andika Risky |
| Issue Date: | 2025 |
| Publisher: | IPB University |
| Abstract: | Persoalan agraria di Indonesia menjadi salah satu persoalan yang perlu
diperhatikan lebih mendalam sebab persoalan agraria erat kaitannya dengan
persengketaan tanah. Berdasarkan Konsorsium Pembaruan Agraria (KPA),
persoalan agraria kembali mengalami kenaikan sejak tahun 2021 hingga 2023.
Banyak faktor yang menyebabkan persoalan agraria terjadi. Salah satu faktor
penyebab tersebut adalah kurangnya pemahaman masyarakat terhadap peraturan
perundang-undangan yang mengatur dasar pokok-pokok agraria. Penyuluhan
kepada masyarakat dengan tujuan meningkatkan pemahaman kini menjadi solusi
yang dijalankan dengan harapan peraturan perundang-undangan yang mengatur
tentang dasar pokok agraria dapat diketahui dan dipahami oleh masyarakat. Namun,
solusi ini menjadi kurang efektif berdasarkan sisi waktu dan biaya. Penyuluh perlu
ke berbagai daerah dengan memakan waktu yang panjang untuk melakukan
penyuluhan. Oleh karena itu, diperlukan pendekatan teknologi informasi yang
menjembatani pengetahuan masyarakat terhadap peraturan perundang-undangan
tentang agraria melalui chatbot yang mampu menjawab pertanyaan masyarakat
seputar peraturan perundang-undangan dasar pokok agraria. Chatbot dibangun
menggunakan Large Language Model dengan pendekatan Retrieval Augmented
Generation (RAG) sehingga jawaban yang diberikan oleh chatbot bersumber dari
peraturan perundang-undangan yang mengatur dasar pokok-pokok agraria. Melalui
penelitian ini harapannya masyarakat mampu mendapatkan informasi terkait
peraturan perundang-undangan tentang dasar pokok-pokok agraria dengan mudah
dan mendorong peningkatan pemahaman masyarakat terhadap peraturan tersebut. Agrarian issues in Indonesia are one of the issues that need to be considered more deeply, because agrarian issues are closely related to land disputes. Based on the Agrarian Reform Consortium (KPA), agrarian issues have increased again from 2021 to 2023. Many factors cause agrarian issues to occur, one of the causes is the lack of public understanding of the laws and regulations governing the basic principles of agrarian law. Counseling the community with the aim of increasing understanding is now a solution that is implemented with the hope that the laws and regulations governing the basic principles of agrarian law can be known and understood by the community. However, this solution is less effective in terms of time and cost. Extension workers need to go to various regions and take a long time to conduct counseling. Therefore, an information technology approach is needed that bridges the gap between laws and regulations on agrarian law and public understanding through chatbots that are able to answer public questions about basic agrarian laws and regulations. The chatbot is built using the Large Language Model with the Retrieval Augmented Generation (RAG) approach so that the answers given by the chatbot are sourced from laws and regulations governing the basic principles of agrarian law. Through this research, it is hoped that the public will be able to obtain information related to laws and regulations on the basic principles of agrarian law easily and encourage increased public understanding of these regulations. |
| URI: | http://repository.ipb.ac.id/handle/123456789/165965 |
| Appears in Collections: | UT - Computer Science |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| cover_G6401211046_f868c6f8402d48a38f41b14d82808661.pdf | Cover | 795.41 kB | Adobe PDF | View/Open |
| fulltext_G6401211046_4590b60cc0cb46e1afcbd880487d8924.pdf Restricted Access | Fulltext | 4.95 MB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.