Syntax : Jurnal Informatika

Comparison Email Spam detection vectorizing using bag of word, TFIDF and Word2Vec in Multinomial Naïve Bayes

2024-05-15T07:21:10+07:00

Email has become very popular among people nowadays. In fact, it the cheapest, popular and fastest means of communication in recent times. Email also has become official communication media in business area. The popularity of email is also used by irresponsible people as a medium for sending fake news, as a medium for fraud and so on. We call this kind email as spam email. There are dangerous and not dangerous spam email. We will focus on detection dangerous spam email, there are 2 type dangerous spam email. The first is email Phishing: Phishing is a term used to define fraudulent practices in which spammers try to trick victims. This can be detrimental to the person who receives these emails. And this kind email may deliver massively and very disturbing the email user. This research will try to find better preprocessing text technique to support the Multinomial Naïve Bayes algorithm with 3 class (ham, phishing and fraud) to classify kind of email, it is hoped that it can help users more accurately classify spam emails. To be able to do that, in preprocessing data we need to vectorizing body email so machine learning can make calculation. Vectorization enables the machines to understand the textual contents by converting them into meaningful numerical representations. The effectiveness of various text vectorization methods, namely the bag of word, TF-IDF and word2vec are investigated for email spam detection using the Multinomial Naïve Bayes. The paper presents the comparative analysis of different vectorization methods on spam email dataset. This paper will give the best vectorization with Multinomial Naive Bayes.

Desain Sistem Peminjaman Peralatan Kantor Berbasis Website

2024-06-13T11:52:19+07:00

Manajemen aset dalam perusahaan menjadi elemen krusial dalam
menjaga kelancaran operasional dan mendukung produktivitas karyawan.
Namun, kekurangan sistem peminjaman yang terstruktur sering kali
mengakibatkan masalah seperti kerusakan, kehilangan aset, kesulitan dalam
pemantauan, dan pencatatan manual yang rentan terhadap kesalahan dan
kehilangan data. Oleh karena itu, tujuan dari penelitian ini adalah untuk
merancang dan membangun sebuah sistem informasi peminjaman aset alat
kantor berbasis website yang dapat menyederhanakan proses peminjaman,
meningkatkan pengawasan terhadap aset kantor, dan menyediakan laporan yang
akurat mengenai peminjaman tersebut. Metode pengembangan sistem yang
digunakan dalam penelitian ini adalah metode waterfall, dan desain sistem
menggunakan UML (Unified Modeling Language). Hasil akhir dari penelitian ini
adalah sistem peminjaman aset alat kantor berbasis website yang menggunakan
framework Laravel, sehingga dapat meningkatkan pengawasan terhadap kondisi
dan ketersediaan aset, mengurangi risiko kehilangan data, dan meningkatkan
akurasi laporan yang dihasilkan.

Penerapan Algoritma Apriori untuk Memprediksi Pembayaran UKT

2024-05-05T14:34:46+07:00

Penelitian ini menerapkan algoritma Apriori untuk memprediksi hasil anilsis pola asosiasi pembayaran cicilan uang kuliah di Universitas Buana Perjuangan Karawang. Aturan asosiasi menunjukkan bahwa pembayaran Cicilan 3 memiliki dampak besar terhadap Cicilan 4, dengan tingkat support sebesar 84.60% dan confidence sebesar 93.47%. Ketergantungan positif antara Cicilan 2 dan Cicilan 3 dengan Cicilan 4 juga teridentifikasi dengan nilai support sebesar 84.57% dan nilai confidence sebesar 94.03%. Rekomendasi kebijakan mencakup penggabungan paket pembayaran pada Cicilan 3 dan Cicilan 4 serta insentif pembayaran lebih awal. Pemodelan menggunakan algoritma Apriori dengan implementasi Python dan Google Colaboratory.

TOPIC MODELING ANALYSIS OF ACCESS BY KAI APPLICATION REVIEWS ON GOOGLE PLAY STORE USING LATENT DIRICHLET ALLOCATION

2024-07-02T14:23:46+07:00

PT Kereta Api Indonesia (KAI) has released a ticket booking application named Access by KAI, which has been downloaded over 10 million times and has received more than 187,000 reviews on Google Playstore. However, with the vast amount of review data for the Access by KAI application, it is still challenging to understand the aspects that need improvement. In this case, topic modeling is necessary to classify the reviews. The aim of this research is to apply the Latent Dirichlet Allocation (LDA) method to model topics of user reviews of the Access by KAI application on Google Playstore and to present recommendations derived from the data dictionary or bag-of-words through a fishbone diagram. This research uses the lifecycle of the data mining methodology, which consists of the stages of problem definition, selecting text data mining approach, data collecting, text standardization, text processing, feature extraction, analysis, and discovery. The results of this research identified a total of 7 topics with a coherence score of 0.40279302. The conclusions from each topic are as follows: Topic 1 discusses application updates, available versions, interface, and the relationship with stations and cities. Topic 2 involves users complaining about decreased application performance after updates. Topic 3 covers the use of the Access by KAI application to book train tickets, highlighting the app version, user experience, and app quality ranging from good to cumbersome. Topic 4 reports user difficulties in accessing, particularly issues with login and payment after app updates. Topic 5 focuses on login difficulties, slow app performance, and issues in the ticket booking and payment process. Topic 6 reflects user disappointment regarding performance decline in speed and login difficulties after updates. Topic 7 addresses user complaints about difficulties in purchasing train tickets through the KAI app following updates or upgrades.

Comparison Of Naïve Bayes And Support Vector Machines In Classifying Sentiment On Twitter About Artificial Intelligence Development

2024-06-24T11:15:36+07:00

Analisis sentimen merupakan bagian dari data mining yang digunakan untuk mengolah dan memproses teks dengan tujuan untuk mengetahui bagaimana opini atau pandangan masyarakat tentang suatu isu atau masalah tertentu. Metode klasifikasi yang digunakan untuk melakukan analisis sentimen pada data berupa teks, diantaranya Naive Bayes dan Support Vector Machine (SVM). Dalam mengevaluasi performa model klasifikasi yang telah dibuat, biasanya akan diukur nilai akurasinya. Oleh karena itu, penelitian ini bertujuan untuk membandingkan performa dari model klasifikasi sentimen yang menggunakan metode Naive Bayes dan SVM, dengan TF-IDF dan CountVectorizer sebagai ekstraksi fitur serta Information Gain sebagai seleksi fitur. Selain itu, digunakan juga N-gram sebagai upaya untuk dapat meningkatkan akurasi model klasifikasi. Penelitian ini menggunakan dataset berupa cuitan pengguna Twitter tentang perkembangan Artificial Intelligence. Data tersebut nantinya dikategorikan menjadi dua kelas, yaitu positif dan negatif, serta akan diolah dengan menggunakan tahapan knowledge discovery in databases (KDD). Hasil penelitian menunjukkan bahwa model hasil Naive Bayes mendapatkan akurasi tertinggi saat menggunakan ekstraksi fitur CountVectorizer, sedangkan model hasil SVM mendapatkan akurasi tertinggi saat menggunakan TF-IDF. Selain itu, penggunaan Information Gain ternyata dapat meningkatkan nilai akurasi model hasil Naive Bayes sebesar 12% menggunakan CountVectorizer dengan N-gram. Namun penggunaan Information Gain justru menurunkan nilai akurasi model hasil SVM sebesar 0,73% menggunakan TF-IDF dengan N-gram.

Analisis Sentimen terhadap Kebijakan Food Estate Menggunakan Algoritma Support Vector Machine

2024-10-07T11:47:58+07:00

The food estate policy has become a key topic in public discussions in Indonesia regarding food security. However, its implementation has sparked reactions on social media, ranging from positive to negative and neutral. This study aims to analyze public sentiment towards the food estate policy using the Support Vector Machine (SVM) algorithm. SVM was chosen for its proven effectiveness in text classification, and previous studies have demonstrated high accuracy in sentiment analysis. Data were collected from the social media platform X using scraping techniques, followed by data processing. The processed data were then classified into three sentiment categories (positive, negative, and neutral) using SVM with linear, RBF, polynomial, and sigmoid kernels. The eval__uation results show that SVM with a linear kernel and parameter C=2 provided the best performance, achieving 79% accuracy, 80% precision, 79% recall, and an F1-score of 79%. These findings indicate that SVM is capable of accurately classifying public sentiment, offering valuable insights for policymakers in eval__uating the social impact of the policy.