Ladya Kusuma, . (2026) IMPLEMENTASI METODE RAG MENGGUNAKAN FRAMEWORK LLAMAINDEX PADA CHATBOT INFORMASI SMAN 28 KABUPATEN TANGERANG. Skripsi thesis, Universitas Pembangunan Nasional Veteran Jakarta.
![[img]](http://repository.upnvj.ac.id/style/images/fileicons/text.png) |
Text
ABSTRAK.pdf Download (11kB) |
|
Text
AWAL.pdf Download (858kB) |
|
Text
BAB 1.pdf Restricted to Repository UPNVJ Only Download (159kB) |
|
Text
BAB 2.pdf Restricted to Repository UPNVJ Only Download (440kB) |
|
Text
BAB 3.pdf Restricted to Repository UPNVJ Only Download (368kB) |
|
Text
BAB 4.pdf Restricted to Repository UPNVJ Only Download (3MB) |
|
Text
BAB 5.pdf Download (261kB) |
|
Text
DAFTAR PUSTAKA.pdf Download (250kB) |
|
Text
DAFTAR RIWAYAT HIDUP.pdf Restricted to Repository UPNVJ Only Download (124kB) |
|
Text
LAMPIRAN.pdf Restricted to Repository UPNVJ Only Download (1MB) |
|
Text
HASIL PLAGIARISME.pdf Restricted to Repository staff only Download (33MB) |
|
Text
ARTIKEL KI.pdf Restricted to Repository staff only Download (450kB) |
Abstract
The information delivery system at SMAN 28 Kabupaten Tangerang still faces challenges in terms of speed, availability, and information consistency, particularly during the New Student Admission Selection (SPMB) period. This study aims to design and optimize a web-based informational chatbot system by implementing the Retrieval-Augmented Generation (RAG) method using the LlamaIndex framework. The development method employed is the prototype approach, with functional testing conducted using Black Box Testing and performance testing using Apache JMeter.System performance optimization is achieved through the implementation of a query caching mechanism using Redis to improve processing efficiency and system response time.Functional testing results indicate that all system features operate according to specifications with a 100% success rate. Performance testing shows that caching reduces the average response time from 18,392 ms to 53 ms in the 30 virtual users scenario and increases throughput from 1.1 requests/second to 3.1 requests/second. The system maintains acceptable performance up to 700 virtual users but begins to experience an error rate of 0.85% at a load of 800 virtual users. This indicates system limitations in handling sudden spike loads under a monolithic architecture and a Docker-based local server environment. The results demonstrate that the implementation of the Retrieval-Augmented Generation method using the LlamaIndex framework is effective for developing a responsive, stable, and reliable school information chatbot.
| Item Type: | Thesis (Skripsi) |
|---|---|
| Additional Information: | [No.Panggil: 2210511108] [Pembimbing 1: Neny Rosmawarni] [Pembimbing 2: Nur Hafifah Matondang] [Penguji 1: Ridwan Raafi’udin] [Penguji 2: Muhammad Panji Muslim] |
| Uncontrolled Keywords: | Chatbot, RAG, LlamaIndex, Caching Query |
| Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science |
| Divisions: | Fakultas Ilmu Komputer > Program Studi Informatika (S1) |
| Depositing User: | LADYA KUSUMA |
| Date Deposited: | 17 Mar 2026 02:56 |
| Last Modified: | 17 Mar 2026 02:56 |
| URI: | http://repository.upnvj.ac.id/id/eprint/42616 |
Actions (login required)
 |
View Item |