Institut Teknologi Telkom Purwokerto

Strengthening on Micro, Small and Medium Enterprises (MSMEs) in Indonesia needs to be supported by the use and good information systems management. Therefore, business actors are expected to be able to use a operational strategy supported by the use of information systems. Workshop is one of the complex MSMEs with integrated warehousing and financial reporting. The problem faced at this workshop UMKM is that there is no synchronization between the existing stock of goods and the sales stock, as well as reporting both warehouse reporting and financial reporting. For this reason, this study aims to build a web-based MSME information system for tire & rims workshops. In building this system, researchers used the agile software development method, namely SCRUM. This method is used because the system can adjust to the needs of the product owner, which is always changing and fast in processing. The results of this scrum stage, namely the product log, sprint backlog, sprint and working increment of the software, can resolve all problems that occur with regard to time, scope and cost issues so that in the implementation of making this system application can reduce the system requirements gap during the sprint process. So that the system can be completed in accordance with the requirements needed by the user. By using this workshop's UMKM information system, all sales operational activities can be monitored properly and sales and financial reports can be well structured.

The progress of information technology is currently growing rapidly. Technology related to the internet is often a solution to most of the problems in existing needs, especially those related to the effectiveness and efficiency of activities and procedures. In this final project, the author discusses websites in the business field, namely e-commerce websites. In this study, the authors chose one of the most popular e-commerce websites in Indonesia this year, namely the shopee website, a website that not only offers products but also puts forward the appearance of the Shopee website which must always be considered because it is the main factor to increase customer purchases. In this study, the quality of the web that will be measured by users, especially for consumers, is based on measuring the quality of the website using the System Usability Scale (SUS). Evaluation of the shopee website is the first step to measure the level of usability on the website. Usability evaluation on the website is carried out to collect opinions from various respondents regarding the functionality of the website. In this study, the results obtained from the calculation of the average usability of the shopee website of 67.08 so that it can be said that the usability of the shopee website on product purchases has entered the OK category.

As part of the early warning system, forest fire detection has a critical role in detecting fire in a forest area to prevent damage to forest ecosystems. In this case, the speed of the detection process is the most critical factor to support a fast response by the authorities. Thus, this article proposes a new framework for fire detection based on combining color-motion-shape features with machine learning technology. The characteristics of the fire are not only red but also from their irregular shape and movement that tends to be constant at specific locations. These characteristics are represented by color probabilities in the segmentation stage, color histograms in the classification stage, and image moments in the verification stage. A frame-based evaluation and an intersection over union (IoU) ratio was applied to evaluate the proposed framework. Frame-based evaluation measures the performance in detecting fires. In contrast, the IoU ratio measures the performance in localizing the fires. The experiment found that the proposed framework produced 89.97% and 10.03% in the true-positive rate and the false-negative rate, respectively, using the VisiFire dataset. Meanwhile, the proposed method can obtain an average of 21.70 FPS in processing time. These results proved that the proposed method is fast in the detection process and can maintain performance accuracy. Thus, the proposed method is suitable and reliable for integrating into the early warning system.

Machine learning and deep learning are subsets of Artificial Intelligence that have revolutionized object detection and classification in images or videos. This technology plays a crucial role in facilitating the transition from conventional to precision agriculture, particularly in the context of weed control. Precision agriculture, which previously relied on manual efforts, has now embraced the use of smart devices for more efficient weed detection. However, several challenges are associated with weed detection, including the visual similarity between weed and crop, occlusion and lighting effects, as well as the need for early-stage weed control. Therefore, this study aimed to provide a comprehensive review of the application of both traditional machine learning and deep learning, as well as the combination of the two methods, for weed detection across different crop fields. The results of this review show the advantages and disadvantages of using machine learning and deep learning. Generally, deep learning produced superior accuracy compared to machine learning under various conditions. Machine learning required the selection of the right combination of features to achieve high accuracy in classifying weed and crop, particularly under conditions consisting of lighting and early growth effects. Moreover, a precise segmentation stage would be required in cases of occlusion. Machine learning had the advantage of achieving real-time processing by producing smaller models than deep learning, thereby eliminating the need for additional GPUs. However, the development of GPU technology is currently rapid, so researchers are more often using deep learning for more accurate weed identification.

The use of the System Usability Scale (SUS) questionnaire in the measurement of e-commerce usability has been widely carried out. However, the SUS score is not an adequate measure to express the level of user acceptance and satisfaction. Other evaluations are needed to complement the usability test, including assessments based on expert judgment. The proposed method consists of two stages, the heuristic evaluation stage, which involves expert judgment, and the SUS questionnaire stage based on user perceptions of the e-commerce website. Input from experts is expected to be able to show better the usability issues faced in using the website. Expert and user perspectives are combined to get user input in design improvements. We collect data from experts and users about their perceptions of the usability of Shoppee e-commerce websites. Most users agree that the Shopee site is excellent (grade B-). The results of the examination by the expert stated that the Shopee site was also excellent. Nine out of ten evaluation criteria scored above 72%. The most usability issue is the flexibility and efficiency of the system, especially problems in search engines.

Imbalanced data classification is challenging task for various datasets in the real world. One of technique to enlarge the sample in minority class is oversampling to fix size as majority class. This research aims to test SMOTE, Borderline-SMOTE, and ADASYN to handle dataset imbalance and to observe its impact toward classification accuracy. Gradient Boosting applied as a classifier and seven datasets are used in this research. Accuracy, recall, precision, F1-Score, AUC were also implemented to measure classifier performance. Experiments showed that oversampling technic increase accuracy from 2% to 11% for the dataset Mammography, Liver Disorders, Diabetes (Pima Indian), Indian Liver, Habberman, and Immunotherapy. Borderline-SMOTE increases higher accuracy compared to other oversampling method. Surprisingly, Breast Cancer Wisconsin has steady accuracy with or without oversampling. Even though, oversampling good for data imbalanced, the sensibility of oversampling algorithm and the nature of dataset must considered.

Most people with diabetes in the world are type 2. We can detect diabetes early to prevent things that are not desirable by checking sugar and insulin levels with the doctor. In addition to using this method, people with diabetes can also be grouped based on data from diabetes examination results. However, most of the data on health examination results have several parameters that are difficult for the public to understand. These problems can be done by means of automatic classification. In addition to these problems, there is another problem in the form of an unbalanced amount of data for diabetics and non-diabetics. This problem can be done by balancing the amount of data using the model to increase the ratio of the amount of data that is small or decrease the ratio of the amount of data that is too much. Purpose: This study aims to detect type 2 diabetes mellitus using the SVM classification model and analyze the results of the comparison using the SMOTE and ADASYN data balancing technique which is the best. Methods/Study design/approach: The research method starts from collecting the diabetes dataset, then the dataset cleaning process is carried out whether there is a null value or not. After applying two oversampling methods to analyze which method is the most appropriate. After the oversampling technique was carried out, data classification was carried out using a support vector machine model to see the accuracy results. Result/Findings: The results obtained by the ADASYN-SVM method are superior to SMOTE-SVM. The ADASYNSVM method has an accuracy of 87.3%, while the SMOTE-SVM has an accuracy of 85.4%. Novelty/Originality/Value: The data used in this study came from the Karya Medika clinic, Indonesia which contains parameters related to type 2 diabetes.

The use of masks on the face in public places is an obligation for everyone because of the Covid-19 pandemic, which claims victims. Indonesia made 3M policies, one of which is to use masks to prevent coronavirus transmission. Currently, several researchers have developed a masked or non-masked face detection system. One of them is using deep learning techniques to classify a masked or non-masked face. Previous research used the MobileNetV2 transfer learning model, which resulted in an F-Measure value below 0.9. Of course, this result made the detection system not accurate enough. In this research, we propose a model with more parameters, namely the DenseNet201 model. The number of parameters of the DenseNet201 model is five times more than that of the MobileNetV2 model. The results obtained from several up to 30 epochs show that the DenseNet201 model produces 99% accuracy when training data. Then, we tested the matching feature on video data, the DenseNet201 model produced an F-Measure value of 0.98, while the MobileNetV2 model only produced an F-measure value of 0.67. These results prove the masked or non-masked face detection system is more accurate using the DenseNet201 model.

Instagram is a social media for sharing images, photos and videos. Instagram has many active users from various circles. In addition to sharing submissions, Instagram users can also give likes and comments to other users' posts. However, the comment feature is often misused, for example it is used for cyberbullying which includes one act against the law. But until now, Instagram still does not provide a feature to detect cyberbullying. Therefore, this study aims to create a system that can classify comments whether they contain elements of cyberbullying or not. The results of the classification will be used to detect cyberbullying comments. The algorithm used for classification is Naïve Bayes Classifier. Then for each comment will pass the preprocessing and feature extraction stages with the TF-IDF method. For evaluation and testing using the K-Fold Cross Validation method. The experiment is divided into two, namely using stemming and without stemming. The training data used is 455 data. The best experimental results obtained an accuracy of 84% both with stemming, and without stemming.

In the vehicular ad hoc network, moving vehicles can keep communicating with each other by entering or leaving the network at any time to establish a new connection. However, since many users transmit a substantial number of messages, it may cause reception delays and affect the entire system. A certificateless aggregate signature scheme can provide a signature compression that keeps the verification cost low. Therefore, it is beneficial for environments constrained by time, bandwidth, and storage, such as vehicular ad hoc network. In recent years, several certificateless aggregate signature schemes have been proposed. Unfortunately, some of them still have some security and privacy issues under specific existing attacks. This article offers an authentication scheme that can improve security, privacy, and efficiency. First, we apply the certificateless aggregate signature method to prevent the onboard unit devices from leaking sensitive information when sending messages. The scheme is proven to be secure against the Type-1 [Formula: see text] and Type-2 [Formula: see text] adversaries in the random oracle model under the computational Diffie–Hellman problem assumption. Then, the performance evaluation demonstrates that our proposed scheme is more suitable for deployment in vehicular ad hoc network environments.

The learning process between students and lecturers usually occurs face-to-face in class. Technological developments and a continuous pandemic change the learning process to be a face-to-face e-learning process. The mental load during face-to-face learning is very different from learning in e-learning. This study was built using ergonomic thinking that is integrated with the use of e-learning. Cognitive ergonomics see from the point of view of students' comfort in cognitive thinking processes when doing e-learning. Data processing and testing will use a questionnaire derived from the NASA-TLX method. The results obtained from this study are the mental load calculations of each NASA TLX calculation. NASA TLX calculations show that efforts with a value of 267.29 dominate students. It could indicate that in e-learning lectures, students need more effort in conducting lectures. In addition, students experience fatigue while participating in online learning. It can be seen from the average SOFI measurement, which is only 1.26.

Akhir tahun 2019 lalu dunia digemparkan oleh munculnya suatu penyakit yang disebabkan oleh virus SARS-CoV-2 yang merupakan jenis virus terbaru dari coronavirus. Penyakit ini dikenal dengan nama COVID-19. Penyebaran penyakit ini terbilang cukup luas dan cepat. Dalam waktu singkat penyakit ini mulai menyebar ke segala penjuru dunia tak terkecuali Indonesia. Dengan tingkat penyebaran yang begitu tinggi dan belum ditemukannya vaksin untuk COVID-19, menyebabkan kekacauan di tengah masyarakat. Hal ini mempengaruhi banyak sektor kehidupan masyarakat. Tak sedikit masyarakat yang aktif bersosial media dan menuliskan pendapat, opini serta pemikirannya di platform media sosial seperti Twitter. Terjadinya pandemi ini mendorong masyarakat untuk menuliskan opini, pemikiran serta pendapatnya terhadap COVID-19 pada media sosial Twitter. Dibutuhkan suatu model sentiment analysis untuk mengklasifikasi tweet masyarakat di Twitter menjadi positif dan negatif. Sentiment analysis merupakan bagian dari Natural Language Processing yang membuat sebuah sistem guna mengenali serta mengekstraksi opini dalam bentuk teks. Pada penelitian ini digunakan algoritma Naive Bayes dan K-Nearest Neighbor untuk digunakan dalam membangun model sentiment analysis terhadap tweet pengguna Twitter terhadap COVID-19. Didapatkan akurasi sebesar 85% untuk algoritma Naïve Bayes dan 82% untuk algoritma K-Nearest Neighbor pada nilai k=6, 8, dan 14.

This research used the 5G NR network planning with the frequency of 3.5 GHz simulated using the Mentum Planet 7.2.1 software with the planning method in the coverage side by employing a case study in Pulogadung industrial zone with a total area of 5 km <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> . This planning used the 5G NR planning method with the propagation model in accordance with the plan, that is, UMa (Urban Macro - Street canyon) model based on the standard of 3GPP TR 38.901, using 8 design scenarios for the uplink; and downlink; outdoor-to-outdoor (O2O) and outdoor-to-indoor (O2I); line of sight (LOS) and non-line of sight (NLOS) conditions. The simulation result showed that the design with the downlink scenario required more sites to fulfill the area services that the uplink scenario since influenced by the link budget parameter, that is, the value of interference margin downlink which is greater than that of the uplink margin interference; From all scenario parameters of SS-RSRP observed, scenario 1 (downlink-O2O-LOS) had the highest average SS-RSRP of -92.95 dBm and the lowest average SS-RSRP resulted from the scenario 2 (uplink-O2O-LOS) of -97.16 dBm. The average SS-RSRP value was influenced by the number of sites covering the planning areas that scenario 1 had the highest SS-RSRP parameter.

Background: The COVID-19 pandemic has made people spend more time on online meetings more than ever. The prolonged time looking at the monitor may cause fatigue, which can subsequently impact the mental and physical health. A fatigue detection system is needed to monitor the Internet users well-being. Previous research related to the fatigue detection system used a fuzzy system, but the accuracy was below 85%. In this research, machine learning is used to improve accuracy.Objective: This research examines the combination of the FaceNet algorithm with either k-nearest neighbor (K-NN) or multiclass support vector machine (SVM) to improve the accuracy.Methods: In this study, we used the UTA-RLDD dataset. The features used for fatigue detection come from the face, so the dataset is segmented using the Haar Cascades method, which is then resized. The feature extraction process uses FaceNet's pre-trained algorithm. The extracted features are classified into three classes—focused, unfocused, and fatigue—using the K-NN or multiclass SVM method.Results: The combination between the FaceNet algorithm and K-NN, with a value of resulted in a better accuracy than the FaceNet algorithm with multiclass SVM with the polynomial kernel (at 94.68% and 89.87% respectively). The processing speed of both combinations of methods has allowed for real-time data processing.Conclusion: This research provides an overview of methods for early fatigue detection while working at the computer so that we can limit staring at the computer screen too long and switch places to maintain the health of our eyes.

This research is expected to be an initial planning for 5G New Radio (NR) technology implementation in Indonesia and discusses 5G NR network planning based on coverage area at frequency 28 GHz in 5 km <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> of Pulogadung industrial estate. Link budget was calculated based on outdoor-to-outdoor (O2O) with Line of Sight (LOS) scenario and Urban Micro (UMi) was used as propagation model standardized by 3GPP TR 38.901. The planning result was simulated using Mentum Planet ver.7.2.1 and resulted in path loss values 110.30 dB for uplink and 109.80 dB for downlink; the cell radius was 214.37 m for uplink and 202.92 m for downlink; it requires 43 sites for uplink scenario and 60 sites for downlink scenario to prepare a good network service. The simulation employed Automatic Site Placement (ASP) to determine site position regardless of the existing data. From the uplink simulation result, SS-RSRP parameter was displayed with minimum value -110.96 dBm; maximum value -68.66 dBm; and average value -99.54 dBm. The downlink simulation result shows SS-RSRP parameter with minimum value -110.96 dBm; maximum value -68.66 dBm; and average value -98.82 dBm.

Perkembangan teknologi saaat ini sangat cepat diberbagai bidang kehidupan manusia. Salah satu penggunaan teknologi adalah di bidang absensi. Absensi merupakan bagian terpenting dalam sebuah instansi, baik instansi pendidikan, kesehatan, perkantoran dan pemerintahan dalam menunjang memonitor kehadiran sehari hari karyawan. Desa Sokaraja Lor merupakan sebuah desa yang terletak di Kecematan Sokaraja Kab. Banyumas Jawa Tengah. Saat ini proses absensi di kantor desa Sokaraja Lor tersebut masih menggunakan Pinjer Print dan juga menggunakan pencatatan menggunakan buku besar. Penggunaan pinjer print ini dinilai kurang efektif karena apa bila tanggan pegawai desa tersebut basah, atau luka maka absensi tidak dapat dilakukan dan ini juga berisiko pegawai desa bisa titip absen kepada pegawai lainnya. Melihat hal itu peneliti merancang sebuah sistem absensi berbasis Face Recognition dengan menggunakan Platform Android. Absensi berbasis Face recognition merupakan absensi yang dilakukan menggunakan deteksi bagian wajah manusia. Kemudian didalam perancangan sistem absensi berbasis face recognition ini peneliti menggunakan pemodelan sistem dengan Undifinied Modeling Language (UML). Dengan dibangunnya sistem absensi ini Desa Sokaraja Lor dapat lebih mudah dalam melakukan absensi dalam setiap kondisi karena sudah berbasis android, kemudian dalam merekap daftar pegawai yang hadir pemarintah desa lebih gampang karena sudah tersimpan dalam sebuah database.

The increasing number of internet users is directly in line with the increasing number of data on the internet that is available for analysis, especially data in text form. The availability of this text data encourages a lot of sentiment analysis research. However, it turns out that the availability of abundant text data is also one of the challenges in sentiment analysis research. Datasets that consist of long and complex text documents require a different approach. In this study, LSTM was chosen to be used as a sentiment classification method. This research uses a movie review dataset that consists of 25,000 review documents, with an average length per review is 233 words. The research uses CBOW and Skip-Gram methods on word2vec to form a vector representation of each word (word vector) in the corpus data. Several dimensions of the word vector was used in this research, there are 50, 60, 100, 150, 200, and 500, this tuning parameter is used to determine their effect on the resulting accuracy. The best accuracy around 88.17% is obtained at the word vector 100 dimension and the lowest accuracy is 85.86% at the word vector 500 dimension.

Sokaraja Lor is a village located in Sokaraja Subdistrict, Banyumas Regency, Central Java Province led directly by the Village Chief. The Village Head is also assisted by the village device consisting of the Village Secretary, Head of Affairs, and Section Head with duties in accordance with their respective fields. As one of the elements of government organizers, the village government has a duty to organize government affairs, community, and village development. Office activities or work related to the management of letters, storage, and documents are called archiving. Recording of incoming and outgoing mail archives at Sokaraja Lor Village Office is still done manually, which is recorded in the ledger and then stored in the storage cabinet. However, stored mail archives are sometimes lost because the borrowed letters are not recorded by the officer. The purpose of this research is to apply the management of mail archives from manual systems into website-based information systems by utilizing information technology to facilitate and assist the work in managing, searching, or inputting. The method used is the waterfall, which is also called a linear sequential model (sequential linear) or classic life cycle. The results of this study are a website system that is expected to improve administrative performance in terms of managing incoming and outgoing mail so that the process of filing letters and mail search becomes easier and faster.

Saat ini merupakan era teknologi informasi global, dimana segala sesuatu dilakukan dengan serba praktis, tepat dan terbaru dengan informasi yang diperoleh dari manapun dan kapanpun. saat ini penjadwalan pelajaran pada SMP NEGERI 31 Padang masih dilakukan secara manual oleh bagian kurikulum, dengan sebelumnya dilakukan rapat pembagian tugas bersama guru mata pelajaran. Dari penentuan banyaknya kelas, banyaknya guru di sekolah dan banyaknya jam mengajar untuk setiap guru masih dilakukan secara manual. Alokasi dan penentuan guru merupakan elemen yang penting dalam penyusunan jadwal mata pelajaran, namun juga menjadi permasalahan yang umum dalam proses penyusunan jadwal. Dengan membangun Sistem Informasi berbasis web dengan Bahasa pemprograman PHP dan MySql serta pemodelan sistem menggunakan UML (Unified Modelling Language) diharapkan mampu memudahkan pengolahan data siswa, guru, dan jadwal mata pelajaran siswa sehingga dapat diperoleh hasil yang cepat, tepat dan akurat pada SMP NEGERI 31 Padang.

YouTube is the most widely used in Indonesia, and it’s reaching 88% of internet users in Indonesia. YouTube’s comments in Indonesian languages produced by users has increased massively, and we can use those datasets to elaborate on the polarization of public opinion on government policies. The main challenge in opinion analysis is preprocessing, especially normalize noise like stop words and slang words. This research aims to contrive several preprocessing model for processing the YouTube commentary dataset, then seeing the effect for the accuracy of the sentiment analysis. The types of preprocessing used include Indonesian text processing standards, deleting stop words and subjects or objects, and changing slang according to the Indonesian Dictionary (KBBI). Four preprocessing scenarios are designed to see the impact of each type of preprocessing toward the accuracy of the model. The investigation uses two features, unigram and combination of unigram-bigram. Count-Vectorizer and TF-IDF-Vectorizer are used to extract valuable features. The experimentation shows the use of unigram better than a combination of unigram and bigram features. The transformation of the slang word to standart word raises the accuracy of the model. Removing the stop words also contributes to increasing accuracy. In conclusion, the combination of preprocessing, which consists of standard preprocessing, stop-words removal, converting of Indonesian slang to common word based on Indonesian Dictionary (KBBI), raises accuracy to almost 3.5% on unigram feature.