Article Per Year (5 Year)
p-Index From 2020 - 2025
1.556
P-Index
This Author published in this journals
All Journal IAES International Journal of Artificial Intelligence (IJ-AI) International Journal of Informatics and Communication Technology (IJ-ICT) CommIT (Communication & Information Technology) JOIV : International Journal on Informatics Visualization ComTech: Computer, Mathematics and Engineering Applications Engineering, Mathematics and Computer Science Journal (EMACS)
Derwin Suhartono
Computer Science Department, Bina Nusantara University
Civil Engineering, Building, Construction & Architecture Computer Science & IT Engineering Industrial & Manufacturing Engineering Mathematics

Published : 15 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 15 Documents
Search

 1   2 
Pengembangan Algoritma Fast Inversion dalam Membentuk Inverted File untuk Text Retrieval dengan Data Skala Besar Suhartono, Derwin
ComTech: Computer, Mathematics and Engineering Applications Vol 3, No 1 (2012): ComTech
Publisher : Bina Nusantara University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21512/comtech.v3i1.2461

Abstract

The rapid development of information systems generates new needs for indexing and retrieval of various kinds of media. The need for documents in the form of multimedia is increasing currently. Thus, the need to store or retrieve now becomes a primary problem. The multimedia type commonly used is text types, as widely seen as the main option in the search engines like Yahoo, Google or others. Essentially, search does not just want to get results, but also a more efficient process. For the purposes of indexing and retrieval, inverted file is used to provide faster results. However, there will be a problem if the making of an inverted file is related to a large amount of data. This study describes an algorithm called Fast Inversion as the development of base inverted file making method to address the needs related to the amount of data.
Single Document Automatic Text Summarization using Term Frequency-Inverse Document Frequency (TF-IDF) Christian, Hans; Agus, Mikhael Pramodana; Suhartono, Derwin
ComTech: Computer, Mathematics and Engineering Applications Vol 7, No 4 (2016): ComTech
Publisher : Bina Nusantara University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21512/comtech.v7i4.3746

Abstract

The increasing availability of online information has triggered an intensive research in the area of automatic text summarization within the Natural Language Processing (NLP). Text summarization reduces the text by removing the less useful information which helps the reader to find the required information quickly. There are many kinds of algorithms that can be used to summarize the text. One of them is TF-IDF (TermFrequency-Inverse Document Frequency). This research aimed to produce an automatic text summarizer implemented with TF-IDF algorithm and to compare it with other various online source of automatic text summarizer. To evaluate the summary produced from each summarizer, The F-Measure as the standard comparison value had been used. The result of this research produces 67% of accuracy with three data samples which are higher compared to the other online summarizers.
Using K-Nearest Neighbor in Optical Character Recognition Ong, Veronica; Suhartono, Derwin
ComTech: Computer, Mathematics and Engineering Applications Vol 7, No 1 (2016): ComTech
Publisher : Bina Nusantara University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21512/comtech.v7i1.2223

Abstract

The growth in computer vision technology has aided society with various kinds of tasks. One of these tasks is the ability of recognizing text contained in an image, or usually referred to as Optical Character Recognition (OCR). There are many kinds of algorithms that can be implemented into an OCR. The K-Nearest Neighbor is one such algorithm. This research aims to find out the process behind the OCR mechanism by using K-Nearest Neighbor algorithm; one of the most influential machine learning algorithms. It also aims to find out how precise the algorithm is in an OCR program. To do that, a simple OCR program to classify alphabets of capital letters is made to produce and compare real results. The result of this research yielded a maximum of 76.9% accuracy with 200 training samples per alphabet. A set of reasons are also given as to why the program is able to reach said level of accuracy.
Aplikasi E-Tour Guide dengan Fitur Pengenalan Image Menggunakan Metode Haar Classifier Suhartono, Derwin; Permana, William Surya; Wiguna, Antoni; Putra, Ferlan Gisman
ComTech: Computer, Mathematics and Engineering Applications Vol 4, No 2 (2013): ComTech
Publisher : Bina Nusantara University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21512/comtech.v4i2.2593

Abstract

Smartphone has became an important instrument in modern society as it is used for entertainment and information searching except for communication. Concerning to this condition, it is needed to develop an application in order to improve smart phone functionality. The objective of this research is to create an application named E-Tour Guide as a tool for helping to plan and manage tourism activity equipped with image recognition feature. Image recognition method used is the Haar Classifier method. The feature is used to recognize historical objects. From the testing result done to 20 images sample, 85% accuracy is achieved for the image recognition feature.
Eve: An Automated Question Answering System for Events Information Christanno, Ivan; Priscilla, Priscilla; Maulana, Jody Johansyah; Suhartono, Derwin; Wongso, Rini
ComTech: Computer, Mathematics and Engineering Applications Vol 8, No 1 (2017): ComTech
Publisher : Bina Nusantara University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21512/comtech.v8i1.3781

Abstract

The objective of this research was to create a closed-domain of automated question answering system specifically for events called Eve. Automated Question Answering System (QAS) is a system that accepts question input in the form of natural language. The question will be processed through modules to finally return the most appropriate answer to the corresponding question instead of returning a full document as an output. Thescope of the events was those which were organized by Students Association of Computer Science (HIMTI) in Bina Nusantara University. It consisted of 3 main modules namely query processing, information retrieval, and information extraction. Meanwhile, the approaches used in this system included question classification, document indexing, named entity recognition and others. For the results, the system can answer 63 questions for word matching technique, and 32 questions for word similarity technique out of 94 questions correctly.
Question Categorization using Lexical Feature in Opini.id Saputra, Christian Eka; Suhartono, Derwin; Wongso, Rini
ComTech: Computer, Mathematics and Engineering Applications Vol 8, No 4 (2017): ComTech
Publisher : Bina Nusantara University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21512/comtech.v8i4.4026

Abstract

This research aimed to categorize questions posted in Opini.id. N-gram and Bag of Concept (BOC) were used as the lexical features. Those were combined with Naïve Bayes, Support Vector Machine (SVM), and J48 Tree as the classification method. The experiments were done by using data from online media portal to categorize questions posted by user. Based on the experiments, the best accuracy is 96,5%. It is obtained by using the combination of Bigram Trigram Keyword (BTK) features with J48 Tree as classifier. Meanwhile, the combination of Unigram Bigram (UB) and Unigram Bigram Keyword (UBK) with attribute selection in WEKA achieves the accuracy of 95,94% by using SVM as the classifier.
Design of Distribution Optimization Application using Firefly Algorithm Manik, Ngarap Imanuel; Nursalim, Yunggih; Suhartono, Derwin
ComTech: Computer, Mathematics and Engineering Applications Vol 8, No 3 (2017): ComTech
Publisher : Bina Nusantara University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21512/comtech.v8i3.2567

Abstract

The goal of this research was to optimize the distribution of goods and computerize. The method consisted of problem identification, analysis, implementation, and evaluation. Firefly algorithm was used as a method for optimizing the distribution of goods. The results achieved are the shortest distribution route of goods in accordance with the existing constraints and low cost of distribution. It can be concluded that the research can beused too ptimizethe distribution of goods and tominimize distributioncost.
Analyzing the Effects of Combining Gradient Conflict Mitigation Methods in Multi-Task Learning Alison, Richard; Jonathan, Welly; Suhartono, Derwin
CommIT (Communication and Information Technology) Journal Vol. 18 No. 1 (2024): CommIT Journal
Publisher : Bina Nusantara University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21512/commit.v18i1.8905

Abstract

Multi-task machine learning approaches involve training a single model on multiple tasks at once to increase performance and efficiency over multiple singletask models trained individually on each task. When such a multi-task model is trained to perform multiple unrelated tasks, performance can degrade significantly since unrelated tasks often have gradients that vary widely in direction. These conflicting gradients may destructively interfere with each other, causing weights learned during the training of some tasks to become unlearned during the training of others. The research selects three existing methods to mitigate this problem: Project Conflicting Gradients (PCGrad), Modulation Module, and Language-Specific Subnetworks (LaSS). It explores how the application of different combinations of these methods affects the performance of a convolutional neural network on a multi-task image classification problem. The image classification problem used as a benchmark utilizes a dataset of 4,503 leaf images to create two separate tasks: the classification of plants and the detection of disease from leaf images. Experiment results on this problem show performance benefits over singular mitigation methods, with a combination of PCGrad and LaSS obtaining a task-averaged F1 score of 0.84686. This combination outperforms individual mitigation approaches by 0.01870, 0.02682, and 0.02434 for PCGrad, Modulation Module, and LaSS, respectively in terms of F1 score.
Utilizing RoBERTa and XLM-RoBERTa pre-trained model for structured sentiment analysis Putri Masaling, Nikita Ananda; Suhartono, Derwin
International Journal of Informatics and Communication Technology (IJ-ICT) Vol 13, No 3: December 2024
Publisher : Institute of Advanced Engineering and Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.11591/ijict.v13i3.pp410-421

Abstract

The surge in internet usage has amplified the trend of expressing sentiments across various platforms, particularly in e-commerce. Traditional sentiment analysis methods, such as aspect-based sentiment analysis (ABSA) and targeted sentiment analysis, fall short in identifying the relationships between opinion tuples. Moreover, conventional machine learning approaches often yield inadequate results. To address these limitations, this study introduces an approach that leverages the attention values of pre-trained RoBERTa and XLM-RoBERTa models for structured sentiment analysis. This method aims to predict all opinion tuples and their relationships collectively, providing a more comprehensive sentiment analysis. The proposed model demonstrates significant improvements over existing techniques, with the XLM-RoBERTa model achieving a notable sentiment graph F1 (SF1) score of 64.6% on the OpeNEREN dataset. Additionally, the RoBERTa model showed satisfactory performance on the multi-perspective question answer (MPQA) and DSUnis datasets, with SF1 scores of 25.3% and 29.9%, respectively, surpassing baseline models. These results underscore the potential of this proposed approach in enhancing sentiment analysis across diverse datasets, making it highly applicable for both academic research and practical applications in various industries.
Machine Learning for Predicting Personality using Facebook-Based Posts Suhartono, Derwin; Ciputri, Marcella Marella; Susilo, Stefanny
Engineering, MAthematics and Computer Science Journal (EMACS) Vol. 6 No. 1 (2024): EMACS
Publisher : Bina Nusantara University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21512/emacsjournal.v6i1.10748

Abstract

Social media contributes a lot to human life. People can share their thoughts through text, photos, and voice through social media. Information from social media can be useful, including in personality research. Personality can generally be known through personality tests. In this research, personality prediction is formed to determine personality through Facebook posts without using a personality test. We create a model based on big five personality traits using 5 machine learning algorithms: Support Vector Machine (SVM), Multinomial Naive Bayes, Decision Tree, K-Nearest Neighbor, and Logistic Regression. Data augmentation was also used for balancing the dataset value and trained using stratified 10-fold cross-validation. This research yields the highest f1 score on Openness using Multinomial Naive Bayes algorithm of 82.31% and the highest average is 68.62%. So the five supervised Machine Learning algorithms used in this research produced Multinomial Naive Bayes as the best algorithm to predict personality based on big five personality traits from user postings on Facebook.
Co-Authors Agus, Mikhael Pramodana Alison, Richard Chandra Halim Christanno, Ivan Christian, Hans Ciputri, Marcella Marella Eka Putra, Nathanael Geordie Fieri, Brillian Jonathan, Welly Karentia, Angel La'la, Joshua Manik, Ngarap Imanuel Maulana, Jody Johansyah Nugroho, Nico Ardian Nursalim, Yunggih Ong, Veronica Permana, William Surya Priscilla, Priscilla Putra, Ferlan Gisman Putri Masaling, Nikita Ananda Rini Wongso Rudy Salim, Calvin Erico Saffachrissa, Maria Astrid Coenradina Saputra, Christian Eka Setiawan, Dennis Susilo, Stefanny Tamara, Shintia Wiguna, Antoni