Claim Missing Document
Check
Articles

Found 16 Documents
Search

Optimasi Algoritma K-Means Clustering dengan Parallel Processing menggunakan Framework R Marieska, Mastura Diana; Lestari, Suci; Mahendra, Calvin; Oktadini, Nabila Rizky; Buchari, Muhammad Ali
JEPIN (Jurnal Edukasi dan Penelitian Informatika) Vol 7, No 1 (2021): Volume 7 No 1
Publisher : Program Studi Informatika

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26418/jp.v7i1.43400

Abstract

Parallel processing sering digunakan untuk melakukan optimasi execution time terhadap algoritma data mining. Pada penelitian ini, parallel processing digunakan untuk melakukan optimasi pada algoritma clustering K-Means. Implementasi algoritma K-means dilakukan dengan memanfaatkan package yang tersedia pada framework R. Algoritma K-Means dijalankan secara serial dan parallel. Untuk mendapatkan persentase optimasi, maka dilakukan perbandingan antara execution time pada parallel processing dan execution time pada serial processing. Penelitian ini menggunakan dataset Boston Housing yang umum digunakan pada data mining. Skenario pengujian dibedakan berdasarkan jumlah core dan jumlah centroid. Hasil pengujian menunjukkan bahwa parallel processing untuk tiap skenario memiliki execution time yang lebih kecil daripada serial processing. Optimasi yang dihasilkan cukup signifikan, yakni bernilai 20% hingga 52%. Optimasi tertinggi didapatkan pada jumlah core terbanyak dan jumlah centroid terbesar.
Pengaruh Synthetic Minority Oversampling Technique pada Analisis Sentimen Menggunakan Algoritma K-Nearest Neighbors Raisha Fatiya; Novi Yusliani; Mastura Diana Marieska; Danny Matthew Saputra
Jurnal Linguistik Komputasional Vol 5 No 1 (2022): Vol. 5, NO. 1
Publisher : Indonesia Association of Computational Linguistics (INACL)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26418/jlk.v5i1.63

Abstract

Salah satu permasalahan yang kerap timbul dan mempengaruhi proses pengklasifikasian adalah data tidak seimbang (imbalanced data). Suatu data dikatakan tidak seimbang apabila data tersebut terbagi menjadi kelas minoritas dan mayoritas. Hal tersebut akan memberikan pengaruh berupa dampak buruk pada hasil klasifikasi karena hasil yang didapatkan akan bias terhadap kelas mayoritas. Penelitian ini bertujuan untuk mengetahui pengaruh Synthetic Minority Oversampling Technique untuk menangani permasalahan imbalance data pada analisis sentimen menggunakan algoritma K-Nearest Neighbors. Synthetic Minority Oversampling Technique (SMOTE) adalah salah satu metode yang dapat digunakan untuk melakukan pembentukan data synthetic untuk mengatasi imbalanced data. Metode klasifikasi yang digunakan adalah K-Nearest Neighbors (KNN) dengan nilai k=3. Penelitian ini menggunakan tiga dataset sentimen yang memiliki topik Covid, Pilkada 1, dan Pilkada 2. Hasil evaluasi dari ketiga dataset tersebut menghasilkan nilai rata-rata accuracy sebesar 70% pada KNN tanpa SMOTE dan menghasilkan 78% pada KNN+SMOTE. Hal ini menunjukkan bahwa SMOTE dapat mengatasi permasalahan imbalanced data dan dapat memberikan pengaruh berupa peningkatan akurasi pada penelitian ini.
Query Reformulation for Indonesian Question Answering System Using Word Embedding of Word2Vec Alvi Syahrini Utami; Novi Yusliani; Mastura Diana Marieska; Abdiansah Abdiansah
Computer Engineering and Applications Journal Vol 11 No 1 (2022)
Publisher : Universitas Sriwijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (553.716 KB) | DOI: 10.18495/comengapp.v11i1.394

Abstract

Query reformulation is one of the tasks in Information Retrieval (IR), which automatically creates new queries based on previous queries. The main challenge of query reformulation is to create a new query whose meaning or context is similar to the old query. Query reformulation can improve the search for relevant documents for Open-domain Question Answering (OpenQA). The more queries are given to the search system, and the more documents will be generated. We propose a Word Predicted and Substituted (WPS) method for query reformulation using a word embedding word2vec. We tested this method on the Indonesian Question Answering System (IQAS). The test results obtained an E-1 value of 81% and an E-2 value of 274%. These results prove that the query reformulation method with WPS and word-embedding can improve the search for potential IQAS answers.
Sosialisasi dan Pelatihan Computational Thinking untuk Guru TK, SD, dan SMP di Sekolah Alam Indonesia (SAI) Palembang Mastura Diana Marieska; Dian Palupi Rini; Nabila Rizky Oktadini; Novi Yusliani; Yunita Yunita
Annual Research Seminar (ARS) Vol 5, No 2 (2019): Special Issue : Pengabdian Kepada Masyarakat
Publisher : Annual Research Seminar (ARS)

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Computational thinking umumnya hanyadipahami oleh kelompok tertentu, terutama orang yangbekerja di bidang informatika. Padahal computationalthinking merupakan skill yang penting untuk dikuasai padaera digital seperti sekarang ini. Di berbagai negara maju,pelajaran wajib yaitu STEM (Science, Technology,Engineering, and Mathematics) telah diperluas menjadiSTEM-C, yaitu penambahan computational thinking sebagaipelajaran wajib di sekolah. Diperlukan sosialisasi yang luasagar masyarakat Indonesia mengenal dan menyadaripentingnya kemampuan computational thinking. Salah satubentuk sosialisasi yang efektif adalah dengan memberipelatihan pada para guru. Pada tanggal 3 November 2018,telah dilakukan sosialisasi dan pelatihan computationalthinking pada guru TK, SD, dan SMP di Sekolah AlamIndonesia Palembang. Pencapaian dari pelatihan ini adalahpara guru memahami lebih dalam mengenai computationalthinking dan memiliki strategi yang nyata untuk menerapkanpembelajaran computational thinking di kelasnya masing-masing.
Sistem Pengamanan Data Menggunakan Kriptografi AES dan Blockchain Berbasis Android Dhiya Calista; Al Farissi; Mastura Diana Marieska
JUPITER (Jurnal Penelitian Ilmu dan Teknologi Komputer) Vol 13 No 2 (2021): JUPITER Edisi Oktober 2021
Publisher : Teknik Komputer Politeknik Negeri Sriwijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.5281/3927.jupiter.2021.10

Abstract

Data or information security is a very important thing for internet users to pay attention to now, so that the data or information owned is not attacked by irresponsible parties. So, in this research, an implementation of a combination of Blockchain and AES cryptography will be carried out in order to avoid active and passive attacks by attackers. Blockchain method can detect data changes from attackers quickly and easily. However, Blockchain method can still be attacked passively, therefore AES method is combined with Blockchain as a complement that is used to encrypt data from plaintext to ciphertext so that existing data or information can be avoided from active or passive attacks. In this research, the software development method is using Rational Unified Process (RUP) method and the tests carried out are Blockchain resistance to modification attacks testing and Avalanche Effect testing on AES method. Keywords— Cryptography, Blockchain, AES, RUP, Avalanche Effect
SISTEM INFORMASI TOPIK TUGAS AKHIR UNTUK MENCEGAH PLAGIARISME DAN KEMIRIPAN TOPIK (Studi Kasus : Program Studi Teknik Informatika Universitas Sriwijaya) Mastura Diana Marieska; Sari Dwi Septiani; Fressy Arlind
Jurnal Sistem Informasi Vol 12, No 2 (2020)
Publisher : Universitas Sriwijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (517.355 KB) | DOI: 10.36706/jsi.v12i2.12326

Abstract

Tugas akhir atau skripsi merupakan komponen penting bagi kelulusan mahasiswa pada program sarjana. Pada hampir semua perguruan tinggi, mahasiswa tingkat akhir disyaratkan untuk melakukan penelitian secara mandiri dibimbing oleh satu atau lebih dosen pembimbing. Tugas akhir dilakukan oleh tiap mahasiswa. Hal ini menimbulkan potensi terjadinya plagiarisme terhadap pengerjaan tugas akhir. Perlu dilakukan usaha agar topik tugas akhir mahasiswa tidak sama persis atau memiliki tingkat kemiripan yang tinggi satu sama lain. Pada penelitian ini, dirancang dan diimplementasikan Sistem Informasi Topik Tugas Akhir Mahasiswa yang dapat diakses secara online. User dari sistem ini adalah mahasiswa, dosen, dan kepala jurusan. User dapat memasukkan keyword untuk melakukan pencarian topik tugas akhir. User juga dapat memilih kategori pencarian. Terdapat 9 kategori pencarian yang diimplementasikan pada Sistem Topik Tugas Akhir, yaitu judul, abstrak, Nomor Induk Mahasiswa (NIM), nama mahasiswa, angkatan, kelas, status, bidang riset, dan dosen pembimbing. Dengan adanya Sistem Topik TA, dosen dapat mengecek kemiripan judul yang diajukan mahasiswa dengan database topik tugas akhir yang pernah dikerjakan mahasiswa sebelumnya, sehingga plagiarisme dapat dihindari.
Optimasi Fuzzy Time Series Chen Pada Prediksi Kasus Covid-19 Di Sumatera Selatan Menggunakan Particle Swarm Optimization HAFIZH SHAFWAN RAFA; Dian Palupi Rini; Mastura Diana Marieska
JUPITER (Jurnal Penelitian Ilmu dan Teknologi Komputer) Vol 14 No 2-c (2022): Jupiter Edisi Oktober 2022
Publisher : Teknik Komputer Politeknik Negeri Sriwijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.5281./4949/5.jupiter.2022.10

Abstract

At the beginning of its appearance, COVID-19 made the whole community become worried about the possibility that would happen in the future. Prediction of COVID-19 cases is a solution that can be done to reduce this worry. This study uses the Fuzzy Time Series Chen method to predict COVID-19 cases in the future, but on the other hand this method has shortcomings in determining the length of the interval which can result in the prediction accuracy being less good, so a Particle Swarm Optimization algorithm is needed to optimize the length. intervals that will later be used to predict cases of COVID-19, so that the results of the predictions will be better. Prediction accuracy is calculated using Mean Absolute Percentage Error. Based on testing the MAPE error value generated from Fuzzy Time Series Chen which is optimized for 26.380%, while for predictions without optimization it produces a value of 30.057%.
Analisis Sentimen di Twitter Menggunakan Algoritma Artificial Neural Network Novi Yusliani; Armenia Yuhafiz; Mastura Diana Marieska; Alvi Syahrini Utami
JUPITER (Jurnal Penelitian Ilmu dan Teknologi Komputer) Vol 15 No 1d (2023): Jupiter Edisi April 2023
Publisher : Teknik Komputer Politeknik Negeri Sriwijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.5281./6603/15.jupiter.2023.04

Abstract

Along with the development of social media, the amount of data in the form of opinions is increasing. The opinions in social media can be used to find out the assessments of social media users regarding something, one of which is the assessment of a candidate in politics. In general, the opinions in social media can be classified into two categories, namely positive and negative. Sentiment analysis is one of the research topics in the field of Natural Language Processing which aims to classify opinions into one of these categories. The opinions in social media that are often used as research objects are the opinions of Twitter users. This study uses an Artificial Neural Network (ANN) algorithm to be implemented in sentiment analysis system. The dataset used in this study is 1088 tweets consisting of 700 tweets labeled positive and 388 tweets labeled negative. The test results show that the best performance is produced when the data is divided into 80% for training and 20% for testing. The resulting percentages for each performance parameter used are accuracy is 61.3%, recall is 67.9%, precision is 75.1%, and f1-score is 71.3% using 0.01 for learning rate and 150 for epoch.
Comparison Of Shift Reduce Parsing and Left Corner Parsing Algorithm in Sentence Structure Ambiguity Checker Reyhan Navind Shaquille; Novi Yusliani; Mastura Diana Marieska
Sriwijaya Journal of Informatics and Applications Vol 2, No 2 (2021)
Publisher : Fakultas Ilmu Komputer Universitas Sriwijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36706/sjia.v2i2.26

Abstract

Indonesian is the official language of the Republic of Indonesia and the language of the Indonesian nation's unity. Although it is often used, there are still errors in the use that are not in accordance with the applicable rules. One type of error is due to ambiguity which can cause misunderstandings in interpreting a word or sentence. Structural ambiguity is a type of ambiguity that occurs when the structure of words in a sentence can be given more than one grammatical structure. Left Corner Parsing and Shift Reduce Parsing are parsing methods used to classify sentence structure ambiguity. This research involves preprocessing, namely case folding, tokenizing and Part Of Speech Tagging. This study uses 90 testing data labeled with facts, 30 ambiguous sentences and 60 unambiguous sentences. Based on the results of checking the ambiguity of the sentence structure, the Shift Reduce Parsing algorithm produces an accuracy of 71%, precision 70.6%, recall 59%, and f-measure 58.2%. Meanwhile, Left Corner Parsing produces an accuracy value of 70%, precision 68.7%, recall 57.5%, and f-measure 55.8%.
The effect of Chi-Square Feature Selection on Question Classification using Multinomial Naïve Bayes Yusliani, Novi; Aruda, Syechky Al Qodrin; Marieska, Mastura Diana; Saputra, Danny Mathew; Abdiansah, Abdiansah
Sinkron : jurnal dan penelitian teknik informatika Vol. 6 No. 4 (2022): Article Research: Volume 6 Number 4, October 2022
Publisher : Politeknik Ganesha Medan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33395/sinkron.v7i4.11788

Abstract

Question classification is one of the essential tasks for question answering system. This task will determine the expected answer type (EAT) of the question given to the system. Multinomial Naïve Bayes algorithm is one of the learning algorithms that can be used to classify questions. At the classification stage, this algorithm used a set of features in the knowledge model. The number of features used can result in curse of dimensionality if the feature is in high dimension. Feature selection can be used to reduce the feature dimension and could increase the system performance. Chi-Square algorithm can be used to select features that describe each category. In this research, the Multinomial Naïve Bayes is used to classify the question sentences and the Chi-Square algorithm is used for the feature selection. The dataset used is a set of Indonesian question sentences, consisting of 519 labeled factoids, 491 labeled non-factoids, and 185 labeled other. The test results showed an increase in accuracy of 0.1 when used feature selection. System accuracy when used feature selection is 0.87 with the number of features used are 248. Without feature selection, the accuracy is 0.77 with the number of features used are 1374.