Suchergebnisse

Penggunaan internet terutama media sosial telah menjadi bagian dari kehidupan bernegara. Hal ini salah satunya karena Anggota Dewan Perwakilan Rakyat Republik Indonesia (DPR RI) banyak yang menyampaikan ide, kebijakan maupun memberikan komentar atas kebijakan pemerintah melalui media sosial. Penelitian ini dilakukan untuk mengukur pendapat atau memisahkan antara sentimen positif dan sentimen negatif terhadap DPR RI. Data yang digunakan dalam penelitian ini didapatkan dengan melakukan crawling pada media sosial twitter. Penelitian dilakukan dengan menggunakan dua Algoritma yaitu Algoritma Support Vector Machine (SVM) dan Naive Bayes (NB). Kedua algoritma tersebut masing-masing dioptimasi menggunakan Particle Swarm Optimization (PSO). Hasil pengujian k-fold cross validation SVM dan NB mendapatkan nilai accuracy 71,04% dan 70,69% dengan nilai Area Under the Curve (AUC) 0,817 dan 0,661. Sedangkan hasil pengujian k-flod cross validation dengan menggunakan PSO, untuk SVM dan NB masing-masing mendapatkan nilai accuracy 75,03% dan 73,49% dengan nilai AUC 0,808 dan 0,719. Penggunaan PSO mampu meningkatkan nilai accuracy algoritma SVM sebesar 3,99% dan 2,8% pada algoritma NB. Hasil dari pengujian kedua algoritma tersebut nilai accuracy tertinggi adalah SVM dengan PSO sebesar 75,03%.

AbstrakPhysical distancing kini sedang ramai menjadi perbincangan publik sebagai salah satu cara pemerintah dalam menekan penyebaran virus covid-19 yang sedang melanda beberapa negara di belahan dunia. Tidak tersaringnya cuitan terkait physical distancing bisa memunculkan berbagai macam opini, tidak hanya opini yang positif tapi juga yang negatif. Maka dari itu, Twitter di anggap lebih diminati oleh masyarakat indonesia dikarenakan twitter dirasa lebih mudah untuk mengungkapkan opininya. Metode yang digunakan yaitu Naive Bayes Classifier (NBC). Data terkumpul dilakukan filter dari cuitan tersebut dengan menghapus data yang double maka setelah di filter data yang terambil yaitu sebanyak 547 tweets. Pada perhitungan analisis sentimen terhadap physical distancing di tengah pandemi covid-19 menggunakan metode NBC memperoleh hasil akurasi sebesar 50,26%. Tujuan dari penelitian ini, agar dapat mengkategorikan opini negatif atau positif, dari pembahasan physical distancing. Nantinya informasi terkait kebijakan Physical Distancing bisa sampai tepat informasinnya kepada masayarakat.Kata Kunci: naïve bayes, physyical distancing, twitter AbstractPhysical distancing is now busy becoming a public conversation as a way for the government to spread the Covid-19 virus which is currently hitting several countries around the world. There are public tweets related to physical distance that is free from various kinds of opinions, not only positive but also negative ones. Therefore, Indonesian people consider Twitter to be more desirable because it is easier for Indonesians to express their opinion. The method used is the Naive Bayes Classifier (NBC). The data collected was filtered from the tweets with double data, then after filtering the data were taken as many as 547 tweets. In calculating the sentiment analysis of physical distance in the middle of the Covid-19 pandemic using the NBC method, it gets an accurate result of 50.26%. The purpose of this study, to find and categorize negative or positive opinions, from the discussion of physical distancing. So that the implementation of the Physical Distance policy can get accurate information to the public.Keywords: naïve bayes, physical distancing, twitter

One of the government's efforts in providing health to the community is the construction of health centers in each sub-district, and the community is expected to be able to take advantage of the health facilities provided by the government. One of the problems that exist in the community is determining the shortest distance to the puskesmas. In Depok City, there are 26 routes that can be passed from the 38 nodes or vertices to the Cilodong Health Center with the starting point of the Depok mayor's office. This study uses a survey research method to calculate the actual distance at each node or vertex, the purpose of this study is to determine the shortest path taken by the starting point from the Depok mayor's office to get to the Cilodong Health Center by applying the dijkstra algorithm. This dijkstra algorithm works by visiting all existing points and making a route if there are 2 routes to the same 1 point then the route that has the lowest weight is chosen so that all points have an optimal route. This quest continues until the final destination point. After doing this research and testing using a simple application to calculate the distance by applying the djikstra algorithm, it was found that the shortest path taken to the destination is through the GDC Main Gate or on the test results in Iteration 26. From the results of this study, people can choose this closest route to save time when viewed from the distance of the existing track. For further research, it is expected to be able to compare two other algorithms and other parameters so that the closest route with the fastest time is obtained.

Indonesia has become one of the countries with a diverse population so that it has the potential to experience social change, one of which is the influence of the media. Media is an information content that is almost a part of human life. One of the impacts of the media is in the health sector, one of which is in determining the Family Planning program. Family planning is one of the Indonesian government programs designed to reduce the speed of population growth. Since the implementation of the Family Planning Program in Indonesia many tools have been used to prevent pregnancy, namely contraception. Selection of a good contraction is certainly one important thing to plan. In determining good kotrasespi certainly there are influences from various things one of which is the media. Measurement of the influence of the media in determining the Family Planning program can be known by applying data mining. Research conducted with data mining uses a standard methodology called the Cross-Industry Center Process for Data Mining (CRISP-DM). The use of decissin tree in this study was done by comparing the same method by looking at the results of three models namely Split Validation, Cross Validation and Decision Tree Split. The results of Split Validation produce an accuracy of 90.50%, Cross Validation produces an accuracy of 91.58% and Decision Tree Split produces an accuracy of 89.83%. The best results are obtained by using cross validation where with the results of research on 1473 records the accuracy value is 91.58% and the AUC value is 0.690, where the results are obtained from the calculation of the True Positive (TP) 1328, False Negative (FN) ) 36, False Positive (FP) is 88 and True Negative (TN) 21. Exposure to the media is said to be good or influential if they do not have children and are Muslim and educate their husbands in junior high school with a low standard of living but the wife has a college education. Keywords: Family Planning, Media Exposure, Data Mining, Decision Tree.

General elections are an important part of the political process so that many political figures participate in the process. Electability is one of the concerns, various things are done to be able to increase the electability of political figures who participate in general elections. Media has become one of the important tools used to increase electability, one of which is online news media. Reader comments can be used as an assessment of political figures in the form of sentiment analysis. However, it is not easy to analyze sentiments from comments on online news media, because comments contain unstructured text, especially in Indonesian text. Text pre-processing in text mining is an important part of getting the basic information contained in the comments. This research uses Indonesian text pre-processing using the Gata Framework Tetmining. Then proceed with extracting information using the Naïve Bayes classification algorithm and Support Vector Machine which are optimized using Particle Swarm Optimization. Tests carried out with both methods get the results that, Particle Swarm Optimization based on Support Vector Machine is the best method with an accuracy of 78.40% and AUC 0.850. This study found an algorithm that was effective in classifying positive and negative comments related to political figures from online news media. ; Pemilihan umum adalah bagian penting dari proses politik sehingga banyak tokoh politik yang berpartisipasi dalam proses tersebut. Elektabilitas adalah salah satu yang menjadi perhatian, berbagai hal dilakukan untuk dapat meningkatkan elektabilitas tokoh politik yang ikut serta dalam pemilihan umum. Media telah menjadi salah satu alat penting yang digunakan untuk meningkatkan elektabilitas, salah satunya adalah media berita online. Komentar dari pembaca dan dapat digunakan sebagai penilaian terhadap tokoh-tokoh politik dalam bentuk analisis sentimen. Namun, tidak mudah untuk menganalisis sentimen dari berbagai komentar di media berita online, karena komentar mengandung teks yang tidak terstruktur, terutama dalam teks Indonesia. Pra-pemrosesan teks dalam penambangan teks merupakan bagian penting untuk mendapatkan informasi dasar yang terkandung dalam komentar. Penelitian ini menggunakan pra-pemrosesan teks Indonesia dengan menggunakan Gata Framework Textmining. Kemudian dilanjutkan dengan mengekstraksi informasi menggunakan metode klasifikasi Naïve Bayes dan Support Vector Machine yang dioptimalkan menggunakan Particle Swarm Optimization. Pengujian yang dilakukan dengan kedua metode mendapatkan hasil bahwa, Support Vector Machine berbasisi Particle Swarm Optimization adalah metode terbaik dengan akurasi 78.40% dan AUC 0.850. Penelitian ini mendapatkan algoritma yang efektif dalam mengklasifikasikan komentar positif dan negatif terkait dengan tokoh-tokoh politik dari media berita online.

The government modernizes the state revenue system by launching the State Revenue Module (MPN), which connects the billing code generation system with the payment system at the collecting agent. This study aims to develop an information system on how to pay state revenues to assist the public in depositing state revenues through payment channels at the collecting agent by applying the Finite State Automata (FSA) modeling concept. The system development uses the waterfall model while collecting the agent's data. The FSA design and testing phase uses the JFLAP application, while the application design stage uses the Laravel framework. System testing uses the black box testing method to determine how far the functioning of the components or menus on the system has come. In addition to making it easier for the public to deposit state revenues, this study also shows that automata theory can help design a payment information application system. This application design offers a website display of information on how to pay state revenues that can run well. Every menu on the system, when used, has no errors so that this system can be utilized by users and developed in applications based on Android and iOS. ; Pemerintah melakukan modernisasi sistem penerimaan negara dengan meluncurkan Modul Penerimaan Negara (MPN) yang menghubungkan sistem pembuatan kode billing dengan sistem pembayaran pada collecting agent. Penelitian ini bertujuan untuk mengembangkan sistem informasi cara bayar penerimaan negara sehingga membantu masyarakat dalam melakukan penyetoran penerimaan negara melalui kanal pembayaran pada collecting agent dengan menerapkan konsep pemodelan Finite State Automata (FSA). Pengembangan sistem menggunakan model waterfall, sedangkan pengumpulan data berasal dari collecting agent. Tahap perancangan dan pengujian FSA dilakukan dengan menggunakan aplikasi JFLAP, sedangkan untuk tahap desain aplikasi menggunakan framework Laravel. Selain untuk memberi kemudahan bagi masyarakat dalam melakukan penyetoran penerimaan ...