Jurnal RESTI (Rekayasa Sistem dan Teknologi Informasi)
Not a member yet
1071 research outputs found
Sort by
Sentiment Analysis of Public Acceptance of Covid-19 Vaccines Types in Indonesia using Naïve Bayes, Support Vector Machine, and Long Short-Term Memory (LSTM)
No full textThe Covid-19 vaccination is a government program during the pandemic to create herd immunity so that people become more productive in their activities. In Indonesia, the Covid-19 vaccination campaign employs a range of vaccines and has sparked a range of responses from the public on social media, particularly Twitter. Users can tweet and communicate with one another on the social networking site Twitter. This study uses a Sentiment Analysis technique using the Nave Bayes (NB), Support Vector Machine (SVM), and Long Short-Term Memory (LSTM) algorithms to conduct a sentiment analysis of public acceptance of the type of Covid-19 vaccine used in Indonesia using Twitter data. Various types of vaccines in Indonesia include Sinovac, Vaksin Covid-19 Bio Farma, AstraZeneca, Pfizer, Moderna, Sinopharm, Novavax, Sputnik-V, Janssen, Convidencia, Zifivax, often confuse the public in determining the objectivity of this opinion. In addition, theoretically, this study also seeks to contrast the NB, SVM, and LSTM algorithms with experimental techniques to obtain the best algorithm model. The stages of the research involved gathering information based on Twitter user opinions about the type of Covid-19 vaccine on Twitter from January 2021 to January 2022. The researcher used Indonesian language tweet data with the keywords #vaksincorona, #vaksincovid19, #vaksinasi, #ayovaksin, #lawancovid19, and #vaksinindonesia. Before modelling, the pre-processing stage consists of case folding, tokenizing, filtering, stemming, and word weighting using TF-IDF. After that, model testing was carried out using Cross Validation with the Python programming language, and evaluation and validation of the test results using the Confusion Matrix. The results showed that the accuracy score of the SVM method for the best model was 84.89%, while for the Naïve Bayes and LSTM algorithms, they were 84.65% and 82.97%, respectively.The Covid-19 vaccination is a government program during the pandemic to create herd immunity so that people become more productive in their activities. In Indonesia, the Covid-19 vaccination campaign employs a range of vaccines and has sparked a range of responses from the public on social media, particularly Twitter. Users can tweet and communicate with one another on the social networking site Twitter. This study uses a Sentiment Analysis technique using the Nave Bayes (NB), Support Vector Machine (SVM), and Long Short-Term Memory (LSTM) algorithms to conduct a sentiment analysis of public acceptance of the type of Covid-19 vaccine used in Indonesia using Twitter data. Various types of vaccines in Indonesia include Sinovac, Vaksin Covid-19 Bio Farma, AstraZeneca, Pfizer, Moderna, Sinopharm, Novavax, Sputnik-V, Janssen, Convidencia, Zifivax, often confuse the public in determining the objectivity of this opinion. In addition, theoretically, this study also seeks to contrast the NB, SVM, and LSTM algorithms with experimental techniques to obtain the best algorithm model. The stages of the research involved gathering information based on Twitter user opinions about the type of Covid-19 vaccine on Twitter from January 2021 to January 2022. The researcher used Indonesian language tweet data with the keywords #vaksincorona, #vaksincovid19, #vaksinasi, #ayovaksin, #lawancovid19, and #vaksinindonesia. Before modelling, the pre-processing stage consists of case folding, tokenizing, filtering, stemming, and word weighting using TF-IDF. After that, model testing was carried out using Cross Validation with the Python programming language, and evaluation and validation of the test results using the Confusion Matrix. The results showed that the accuracy score of the SVM method for the best model was 84.89%, while for the Naïve Bayes and LSTM algorithms, they were 84.65% and 82.97%, respectively
Aspect Based Sentiment Analysis Marketplace Product Reviews Using BERT, LSTM, and CNN
Get PDFBukalapak is one of the largest marketplaces in Indonesia. Reviews on Bukalapak are only in the form of text, images, videos, and stars without any special filters. Reading and analyzing manually makes it difficult for potential buyers. To help with this, we can extract this review by using aspect-based sentiment analysis because an entity cannot be represented by just one sentiment. Several previous research stated that using LSTM-CNN got better results than using LSTM or CNN. In addition, using BERT as word embedding gets better results than using word2vec or glove. For this reason, this study aims to classify aspect-based sentiment analysis from the Bukalapak marketplace with BERT as word embedding and using the LSTM-CNN method, where LSTM is for aspect extraction and CNN for sentiment extraction. Based on testing the LSTM-CNN method, it gets better results than LSTM or CNN. The LSTM-CNN model gets an accuracy of 93.91%. Unbalanced dataset distribution can affect model performance. With the increasing number of datasets used, the accuracy of a model will increase. Classification without using stemming on datasets can increase accuracy by 2.04%.
Bukalapak is one of the largest marketplaces in Indonesia. Reviews on Bukalapak are only in the form of text, images, videos, and stars without any special filters. Reading and analyzing manually makes it difficult for potential buyers. To help with this, we can extract this review by using aspect-based sentiment analysis because an entity cannot be represented by just one sentiment. Several previous research stated that using LSTM-CNN got better results than using LSTM or CNN. In addition, using BERT as word embedding gets better results than using word2vec or glove. For this reason, this study aims to classify aspect-based sentiment analysis from the Bukalapak marketplace with BERT as word embedding and using the LSTM-CNN method, where LSTM is for aspect extraction and CNN for sentiment extraction. Based on testing the LSTM-CNN method, it gets better results than LSTM or CNN. The LSTM-CNN model gets an accuracy of 93.91%. Unbalanced dataset distribution can affect model performance. With the increasing number of datasets used, the accuracy of a model will increase. Classification without using stemming on datasets can increase accuracy by 2.04%
Sentiment Analysis on Social Media with Glove Using Combination CNN and RoBERTa
No full textTwitter is a popular social media platform that allows users to share short message’s opinion and engage in real-time conversations on a wide range of topics known as tweet. However, tweets often have a complicated and unclear context, which makes it difficult to determine the actual emotion. Therefore, sentiment analysis is required to see the tendency of an opinion, whether the opinion tends to be positive, negative, or neutral. Researchers or institutions can find out how the response and emotions of an issue are happening and make good decisions. With the large user of Twitter social media in Indonesia, sentiment analysis will be carried out using deep learning Convolutional Neural Network (CNN), Term Frequency-Inverse Document Frequency (TF-IDF), Robustly Optimized BERT Pretraining Approach (RoBERTa), Synthetic Minority Over-sampling Technique (SMOTE), and Global Vector (Glove). In this research, the dataset used is trending topics with hashtags related to government policies on Twitter social media and obtained through crawling. By using 30.811 data, the result shows the highest accuracy of 95.56% using CNN with a split ratio of 90:10, baseline unigram, RoBERTa, SMOTE, and Top10 corpus tweet with an increase 10.1%.Twitter is a popular social media platform that allows users to share short message’s opinion and engage in real-time conversations on a wide range of topics known as tweet. However, tweets often have a complicated and unclear context, which makes it difficult to determine the actual emotion. Therefore, sentiment analysis is required to see the tendency of an opinion, whether the opinion tends to be positive, negative, or neutral. Researchers or institutions can find out how the response and emotions of an issue are happening and make good decisions. With the large user of Twitter social media in Indonesia, sentiment analysis will be carried out using deep learning Convolutional Neural Network (CNN), Term Frequency-Inverse Document Frequency (TF-IDF), Robustly Optimized BERT Pretraining Approach (RoBERTa), Synthetic Minority Over-sampling Technique (SMOTE), and Global Vector (Glove). In this research, the dataset used is trending topics with hashtags related to government policies on Twitter social media and obtained through crawling. By using 30.811 data, the result shows the highest accuracy of 95.56% using CNN with a split ratio of 90:10, baseline unigram, RoBERTa, SMOTE, and Top10 corpus tweet with an increase 10.1%
Deteksi Risiko Kredit dalam Peer-to-Peer Lending Menggunakan CatBoost
No full textP2P lending (Peer-to-peer lending) is widely used by private borrowers, small businesses, and MSMEs because P2P lending allows individuals and businesses to be able to lend money directly from lenders without the stringent requirements and criteria of traditional banks and financial institutions. However, P2P lending has a credit risk problem characterized by a high failure rate for borrowers to repay their loans. Therefore, a system was necessary to detect credit risk to minimize the risk of P2P lending. In this study, a system had been built using the CatBoost method; the dataset used was taken from the Bondora loan dataset. To measure the performance of the CatBoost algorithm, an evaluation matrix was performed using ROC (Receiver Operating Characteristics) curves and AUC (Area Under Curve) was performed. The experiment consists of three scenarios, of which the best result regards Scenario 2 with a data splitting of 90:10. It was caused by the result of AUC value 0.80329 compared to scenario 1 with a data split of 80:20 with the AUC value around 0.789583, and scenario 3 with a data split of 70:30 with the AUC value around 0.781066, respectively
Modelling Metadata and Data from Censuses and Surveys with Graph Databases
No full textRelational database users are switching to non-relational databases because non-relational databases are better able to handle dynamic data storage. One of the institutions that require dynamic data storage is Statistics Indonesia (BPS). Currently, data storage for census and survey activities at BPS is done using a relational database, although there are metadata changes in each activity. Accommodating metadata changes in each activity requires one database, which creates problems when retrieving some raw data. There is an opportunity for convenience if the data collected is stored in a non-relational database, one of which is a graph database. This research discusses the modeling of metadata and data from censuses and surveys at BPS using a graph database. Then we implement the Neo4j DBMS and compare the proposed model with the relational model in the Microsoft SQL Server DBMS. Then, a comparison of the features and characteristics of each DBMS is done, and finally, performance testing is done with Apache JMeter. Modeling has been able to handle dynamic data structure changes, but Neo4j's performance is still lagging behind Microsoft SQL Server.Pengguna database relasional beralih ke database non-relasional karena database non-relasional lebih mampu menangani penyimpanan data yang dinamis. Salah satu lembaga yang membutuhkan penyimpanan data dinamis adalah Badan Pusat Statistik (BPS). Saat ini penyimpanan data kegiatan sensus dan survei di BPS dilakukan dengan menggunakan database relasional, padahal terjadi perubahan metadata dalam tiap kegiatannya. Untuk menampung perubahan metadata pada tiap kegiatan membutuhkan satu database, yang menimbulkan permasalahan saat pengambilan beberapa raw data. Ada peluang kemudahan jika data yang dikumpulkan disimpan dalam database non-relasional salah satunya graph database. Penelitian ini membahas pemodelan metadata dan data dari sensus dan survei di BPS menggunakan graph database. Dilanjutkan dengan implementasi pada DBMS Neo4j serta membandingkan model usulan dengan model relasional pada DBMS Microsoft SQL Server. Kemudian dilakukan perbandingan fitur dan karakteristik masing-masing DBMS, dan yang terakhir dilakukan pengujian performa dengan Apache Jmeter. Pemodelan sudah dapat menngani perubahan struktur data dinamis namun performa Neo4j masih tertinggal dibandingkan Microsoft SQL Server
Analysis of the Saintekmu Website Quality on User Satisfaction Using the Modified System Usability Scale and Webqual 4.0 Method
Get PDFToday, websites are a major means of finding or providing information. The Saintekmu website was created to offer top-notch service to students. One way to ensure that the website's services are appropriate and that information technology is being used to its fullest potential is to evaluate the level of service provided and improve its quality. This study aims to compare the results of two methods - the System Usability Scale (SUS) and Webqual - used to determine the quality and expectations of website users. The study distributed questionnaires online using Google Forms and had a sample size of 20 students. The data collected was analyzed using the SPSS program. The results of the SUS method indicated that the website acceptability range was in the marginal category, with a score of 69.9 and a classification rating of OK. The Webqual method yielded an R square of 0.948, indicating that the website's usability, quality, and interaction variables had a significant effect on user satisfaction. All WebQual 4.0 dimensions had a positive and significant effect on user satisfaction, both partially and simultaneously. This study provides Muhammadiyah Saintek University with reference material to evaluate its website in the future
Machine Learning Analisis Klasifikasi dalam Penentuan Status Gizi Anak
No full textMalnutrition is one of the problems that occurs in children due to a lack of nutritional intake. Indonesia contributed 36%, making it the fifth country with the largest cases of malnutrition in the world. On this basis, a solution is needed to reduce the growth rate of malnutrition cases. This research aims to carry out classification analysis to determine nutritional status by optimizing machine learning (ML) performance. The ML classification analysis process will later utilize the performance of the artificial neural network (ANN) method with the Multilayer Perceptron (MLP) algorithm. ML performance can be optimized using the Pearson’s correlation (PC) method to produce optimal classification analysis patterns. This research data set uses child nutrition case data from 576 patients sourced from the M. Djamil Padang Province Regional General Hospital (RSUP). The data set is divided into 417 training data and 159 test data. On the basis of the tests that have been carried out, the performance of the PC method can provide precise and accurate analysis patterns. This analysis pattern has also been able to provide a fairly good level of accuracy, namely 95%. Not only that, this research is also able to present analysis patterns with the best ANN architectural model in classifying nutritional status. Based on the overall results, this research can be used as an alternative solution to the treatment of nutritional problems in children.Gizi buruk merupakan salah satu permasalahan yang terjadi pada kalangan anak-anak yang diakibatkan oleh faktor kurangnya asupan gizi. Fakta yang terjadi bahwa Indonesia telah menyumbang angka 36% dengan menjadikan negara peringkat ke lima dengan kasus malnutrisi terbesar di dunia. Berdasarkan hal tersebut maka dibutuhkan sebuah solusi dalam menekan laju angka pertumbuhan kasus gizi buruk. Penelitian ini bertujuan untuk melakukan analisis klasifikasi penentuan status gizi dengan menggunakan Artificial Neural Network (ANN). Proses analisis klasifikasi ANN nantinya akan memanfaatkan performa kinerja algoritma Multilayer Perceptron (MLP) yang di optimalkan dengan metode Pearson Correlation (PC) untuk menghasilkan keluaran optimal. Pada dasarnya metode PC mampu memberikan peran aktif dalam mengukur kinerja analisis Machine Learning (ML). Adapun dataset penelitian ini menggunakan data kasus gizi anak yang terjadi pada periode tahun sebelumnya. Dataset tersebut bersumber dari Rumah Sakit Umum Propinsi (RSUP) M. Djamil Padang. Berdasarkan pengujian yang telah dilakukan bahwa kinerja metode PC pada proses klasifikasi ANN mampu menyajikan pola analisis yang tepat dan akurat. Pola analisis tersebut juga telah mampu memberikan tingkat akurasi yang cukup baik sebesar 95%. Tidak hanya itu, penelitian ini juga mampu menyajikan pola analisis dengan model arsitektur ANN terbaik dalam klasifikasi status gizi. Berdasarkan keseluruhan hasil, maka penelitian ini dapat dijadikan sebuah solusi alternatif dalam penanganan masalah kasus gizi pada anak. Dengan hasil tersebut maka penelitian ini juga berkontribusi bagi Dinas Kesehatan Provinsi Sumatera Barat dalam menekan angka kasus gizi buruk yang terjadi pada periode tahun berikutnya
Bahasa Inggris
No full textFor the optimization of computer networks with high bandwidth requirements, it is necessary to predict the traffic of the wireless network. Its goal is to reduce maintenance costs and improve internet services. Feature selection is a major issue in multivariate time series (MTS) spatio-temporal modeling. Another problem is the dependency between input features, time lags, and spatial factors, so an appropriate model is needed. This study aims to provide solutions to two problems. The first is to improve a feature extraction and selection process in spatio-temporal MTS data for relevant features using Detrended Partial Cross-Correlation Analysis (DPPCA) and nonredundant features associated with linear using Pearson's correlation (PC) filters and non-linear associations using Symmetrical Uncertainty (SU) and a combination of both PCSUF. The second is to develop a spatiotemporal framework model using recurrent neural networks (RNNs) to get better performance than the traditional model. These methods are combined and tested using a data set of cellular networks with one hour intervals during November in three locations. Testing the effectiveness of the feature selection technique showed that 27.6% of the total extracted features were. The forecasting model with the DPCCA-SU-RNN combination method is the best performance by having RMSE = 380.7, R2 = 97% and MAPE = 10%
Development of Quantum Circuit Architecture on Quantum Perceptron Algorithm for Classification of Marketing Bank Data
Get PDFThe creation of quantum circuit architecture based on the quantum perceptron algorithm to classify marketing bank data is proposed in this work. A quantum circuit is a quantum gate made up of two quantum gates. Quantum bits are used in this study's computation. The primary proposed learning method was not ideal, which is the context of this study. The percentage of qubits measurement value is still 90.7 percent. It is essential to raise the value of the qubit rate. Using the IBM Quantum Experience quantum computer, researchers measured, trained, and tested the quantum circuit architecture. Bank marketing data from the UCI Machine Learning Repository was used. A quantum circuit architecture model results from this research the quantum circuit measurement results.The creation of quantum circuit architecture based on the quantum perceptron algorithm to classify marketing bank data is proposed in this work. A quantum circuit is a quantum gate made up of two quantum gates. Quantum bits are used in this study's computation. The primary proposed learning method was not ideal, which is the context of this study. The percentage of qubits measurement value is still 90.7 percent. It is essential to raise the value of the qubit rate. Using the IBM Quantum Experience quantum computer, researchers measured, trained, and tested the quantum circuit architecture. Bank marketing data from the UCI Machine Learning Repository was used. A quantum circuit architecture model results from this research the quantum circuit measurement results
Web-based CNN Application for Arabica Coffee Leaf Disease Prediction in Smart Agriculture
Get PDFIn the agriculture industry, plant diseases provide difficulty, particularly for Arabica coffee production. A first step in eliminating and treating infections to avoid crop damage is recognizing ailments on Arabica coffee leaves. Convolutional neural networks (CNN) are rapidly advancing, making it possible to diagnose Arabica coffee leaf damage without a specialist's help. CNN is aimed to find features adaptively through backpropagation by adding layers including convolutional layers and pooling layers. This study aims to optimize and increase the accuracy of Arabica coffee leaf disease classification utilizing the neural network architectures: ResNet50, InceptionResNetV4, MobileNetV2, and DensNet169. Additionally, this research presents an interactive web platform integrated with the Arabica coffee leaf disease prediction system. Inside this research, 5000 image data points will be divided into five classes—Phoma, Rust, Cescospora, healthy, and Miner—to assess the efficacy of CNN architecture in classifying images of Arabica coffee leaf disease. 80:10:10 is the ratio between training data, validation, and testing. In the testing findings, the InceptionResnetV2 and DensNet169 designs had the highest accuracy, at 100%, followed by the MobileNetV2 architecture at 99% and the ResNet50 architecture at 59%. Even though MobileNetV2 is not more accurate than InceptionResnetV2 and DensNet169, MobileNetV2 is the smallest of the three models. The MobileNetV2 paradigm was chosen for web application development. The system accurately identified and advised treatment for Arabica coffee leaf diseases, as shown by the system's implementation outcomes.
In the agriculture industry, plant diseases provide difficulty, particularly for Arabica coffee production. A first step in eliminating and treating infections to avoid crop damage is recognizing ailments on Arabica coffee leaves. Convolutional neural networks (CNN) are rapidly advancing, making it possible to diagnose Arabica coffee leaf damage without a specialist's help. CNN is aimed to find features adaptively through backpropagation by adding layers including convolutional layers and pooling layers. This study aims to optimize and increase the accuracy of Arabica coffee leaf disease classification utilizing the neural network architectures: ResNet50, InceptionResNetV4, MobileNetV2, and DensNet169. Additionally, this research presents an interactive web platform integrated with the Arabica coffee leaf disease prediction system. Inside this research, 5000 image data points will be divided into five classes—Phoma, Rust, Cescospora, healthy, and Miner—to assess the efficacy of CNN architecture in classifying images of Arabica coffee leaf disease. 80:10:10 is the ratio between training data, validation, and testing. In the testing findings, the InceptionResnetV2 and DensNet169 designs had the highest accuracy, at 100%, followed by the MobileNetV2 architecture at 99% and the ResNet50 architecture at 59%. Even though MobileNetV2 is not more accurate than InceptionResnetV2 and DensNet169, MobileNetV2 is the smallest of the three models. The MobileNetV2 paradigm was chosen for web application development. The system accurately identified and advised treatment for Arabica coffee leaf diseases, as shown by the system's implementation outcomes.