Orkestrasi Pipeline Machine Learning Berbasis Low-Code Untuk Analisis Sentimen Pada Berita Saham Indeks IDXTECHNO

Suryanto, Rama Prawira (2026) Orkestrasi Pipeline Machine Learning Berbasis Low-Code Untuk Analisis Sentimen Pada Berita Saham Indeks IDXTECHNO. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5024221063-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (2MB) | Request a copy

Abstract

Pasar saham, khususnya indeks IDXTECHNO, sangat dipengaruhi oleh sentimen berita digital. Namun, penerapan analisis sentimen yang otomatis, andal, dan terintegrasi masih menjadi tantangan. Penelitian ini bertujuan merancang low-code untuk analisis sentimen berita saham. Pendekatan rekayasa sistem diterapkan dengan membangun arsitektur machine learning pipeline pada platform n8n menggunakan queue mode serta mengintegrasikan model IndoBERT melalui API REST. Sistem memproses data dari berbagai RSS Feed berita secara otomatis. Hasil pengujian menunjukkan bahwa sistem memiliki reliabilitas operasional yang baik dan mampu meningkatkan skalabilitas melalui penambahan worker. Temuan ini menunjukkan bahwa n8n dapat digunakan sebagai solusi machine learning pipeline yang efisien untuk otomasi analisis sentimen berita saham. Meskipun demikian, proses akuisisi data masih menjadi komponen dengan kontribusi latensi terbesar sehingga perlu dioptimalkan pada pengembangan selanjutnya.
=================================================================================================================================
Digital news sentiment plays an important role in shaping stock market movements, particularly in the IDXTECHNO index. Despite the growing use of sentiment analysis, building an automated, reliable, and integrated pipeline for processing stock-related news remains a practical challenge. This research designs a low-code, end-to-end machine learning pipeline for stock news sentiment analysis. Using a systems engineering approach, the pipeline is implemented on the n8n platform with queue mode and integrated with an IndoBERT-based sentiment model through a REST API. The system automatically collects and processes news from multiple RSS feeds. The evaluation results indicate that the system achieves reliable operational performance and scales more effectively as additional workers are deployed. These findings show that n8n can serve as an efficient platform for automating machine learning pipelines in stock news sentiment analysis. Nevertheless, data acquisition remains the main latency bottleneck and should be further optimized in future development.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	analisis sentimen, IDXTECHNO, IndoBERT, machine learning pipeline, n8n,sentiment analysis, IDXTECHNO, IndoBERT, machine learning pipeline, n8n.
Subjects:	T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T57.74 Linear programming T Technology > T Technology (General) > T58.5 Information technology. IT--Auditing
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Computer Engineering > 90243-(S1) Undergraduate Thesis
Depositing User:	Rama Prawira Suryanto
Date Deposited:	21 Jul 2026 02:20
Last Modified:	21 Jul 2026 02:20
URI:	http://repository.its.ac.id/id/eprint/134954

Actions (login required)

View Item