Garuda - Garba Rujukan Digital

Analisis Sentimen dan Emosi Vaksin Sinovac pada Twitter menggunakan NaÃ¯ve Bayes dan Valence Shifter Akbar, Bagus Muhammad; Akbar, Ahmad Taufiq; Husaini, Rochmat
Jurnal Teknologi Terpadu Vol 7 No 2: Desember, 2021
Publisher : LPPM STT Terpadu Nurul Fikri

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54914/jtt.v7i2.433

The Sinovac vaccine is among the Covid-19 news in the world in early 2021. That information has led to public responses between the pros and cons. Through Twitter media, the public responds to the issue of the Sinovac; therefore, their opinions on Twitter can analyze to count the percentage of sentiment and emotion towards the Sinovac. This analysis hopes to provide a wise and objective reference, although the pros and cons information is still emerging. This study uses Rstudio for sentiment analysis through Twitter opinion classification using NaÃ¯ve Bayes and the Valence Shifter Lexicon method to analyze emotions, also using the NaÃ¯ve Bayes. The Data is 2000 English-language Twitter comments limited to the latest and most popular tweet based on the Sinovac keyword since February 1, 2021, from all Twitter users worldwide. The results showed that NaÃ¯ve Bayes recognized 1433 (71.65%) positive sentiments, 403 (20.15%) negative sentiments, and 164 (8.2%) neutral sentiments. Meanwhile, Valence Shifter Lexicon recognized 903 (45.15%) positive sentiment, 437 (21.85%) negative sentiment, and 660 (33%) neutral sentiments. The NaÃ¯ve Bayes also succeeded in recognizing emotions with the highest number 1727 (86.35%) mixed emotions and 141 (7.05%) joy emotion.

Performance Analysis of FastAPI Framework on Lost Circulation Handling Management Application in Oil Well Drilling Suryotomo, Andiko Putro; Akbar, Bagus Muhammad; Husaini, Rochmat
Telematika Vol 21, No 1 (2024): Edisi Februari 2024
Publisher : Jurusan Informatika

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31315/telematika.v21i1.13259

Purpose: This study aims to conduct a load testing using JMeter and then analyze the performance of the FastAPI framework on the backend of the lost circulation handling management application in oil well drilling. The developed API receives input in the form of drilling parameter data (daily drilling report) from drilling engineers to be processed by a machine learning model (prediction and classification) through the FastAPI framework. The developed API returns processing data in JSON format.Methodology: Performance measurement is done by conducting load testing simulations using the help of JMeter software. Load testing scenarios are created by varying the number of users and ramp-up time, as well as the method of loading the machine learning model used (normal or preemptive loading). The parameters measured in the test scenario are average execution time, maximum execution time, error percentage, and request throughput.Findings: Load testing on a FastAPI-developed API demonstrated that for compute-heavy tasks like machine learning inference, increasing the number of processor cores and using preemptive model loading led to significantly better performance improvements than changes in processor clock speed or switching from HDD to SSD. Even when simulating a higher user load than initially expected (up to 250 users/threads), FastAPI maintained good response times and a low error rate, remaining below 20%.Originality/value/state of the art: This study result is an information about the performance of the FastAPI framework in the application of lost circulation handling management in oil well drilling in the deployment phase, not only up to the model testing phase as in previous studies.

Enhancing Sentiment and Emotion Classification with LSTM-Based Semi-Supervised Learning Husaini, Rochmat; Cahyana, Nur Heri; Wisnalmawati, Wisnalmawati; Mardiana, Tri; Fauziah, Yuli
Compiler Vol 14, No 1 (2025): May
Publisher : Institut Teknologi Dirgantara Adisutjipto

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.28989/compiler.v14i1.2965

The evolution of sentiment analysis has increasingly relied on semi-supervised learning (SSL) models, particularly due to their efficiency in utilizing large amounts of unlabeled data. This study employed four Indonesian datasets—Ridife (sentiment classification), Emotion Indonlu (emotion classification), Sentiment Indonlu (sentiment classification), and Hate Speech (offensive content detection). The LSTM model was trained using labeled data and used to generate pseudo-labels for unlabeled data across three iterations. The performance of the pseudo-labels was evaluated using Random Forest, Logistic Regression, and Support Vector Machine (SVM). The LSTM model demonstrated varying effectiveness across different datasets. For the Sentiment Ridife dataset, LSTM achieved an accuracy of 70.23%, slightly lower than Random Forest but higher than Logistic Regression and SVM. In the Sentiment IndoNLU dataset, LSTM's accuracy was 86.12%, showing strong performance but slightly below Random Forest and Logistic Regression. The Emotion IndoNLU dataset revealed similar performance across models, while the Hate Speech dataset saw LSTM perform well with an accuracy of 86.49%. The results indicate that while LSTM-based SSL can effectively generate pseudo-labels and enhance model performance, its performance varies depending on the dataset and task. This study underscores the need for further research into optimizing pseudo-labeling techniques and exploring advanced NLP models to improve sentiment and emotion analysis in diverse languages.

Performance Analysis of FastAPI Framework on Lost Circulation Handling Management Application in Oil Well Drilling Suryotomo, Andiko Putro; Akbar, Bagus Muhammad; Husaini, Rochmat
Telematika Vol 21 No 1 (2024): Telematika : Jurnal Informatika dan Teknologi Informasi
Publisher : Jurusan Informatika

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31315/telematika.v21i1.13259

Purpose: This study aims to conduct a load testing using JMeter and then analyze the performance of the FastAPI framework on the backend of the lost circulation handling management application in oil well drilling. The developed API receives input in the form of drilling parameter data (daily drilling report) from drilling engineers to be processed by a machine learning model (prediction and classification) through the FastAPI framework. The developed API returns processing data in JSON format.Methodology: Performance measurement is done by conducting load testing simulations using the help of JMeter software. Load testing scenarios are created by varying the number of users and ramp-up time, as well as the method of loading the machine learning model used (normal or preemptive loading). The parameters measured in the test scenario are average execution time, maximum execution time, error percentage, and request throughput.Findings: Load testing on a FastAPI-developed API demonstrated that for compute-heavy tasks like machine learning inference, increasing the number of processor cores and using preemptive model loading led to significantly better performance improvements than changes in processor clock speed or switching from HDD to SSD. Even when simulating a higher user load than initially expected (up to 250 users/threads), FastAPI maintained good response times and a low error rate, remaining below 20%.Originality/value/state of the art: This study result is an information about the performance of the FastAPI framework in the application of lost circulation handling management in oil well drilling in the deployment phase, not only up to the model testing phase as in previous studies.

Title

Found 4 Documents
Search

Abstract

Abstract

Abstract

Abstract

Title Search

Found 4 Documents Search

Abstract

Abstract

Abstract

Abstract

Title

Found 4 Documents
Search