Claim Missing Document
Check
Articles

Found 35 Documents
Search

Implementasi Rekomendasi Content Based Filtering dan Apriori Berbasis Android Mardani, Latif Dwi; Gunawan, Wawan
JEPIN (Jurnal Edukasi dan Penelitian Informatika) Vol 10, No 2 (2024): Volume 10 No 2
Publisher : Program Studi Informatika

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26418/jp.v10i2.74383

Abstract

Proban Motoparts merupakan sebuah toko retail dan jasa yang bergerak dibidang otomotif dengan menjual beberapa produk suku cadang untuk sepeda motor. Banyaknya variasi produk yang ada di Proban Motoparts, membuat pelanggan merasa kesulitan saat memilih produk yang dibutuhkan. Solusi dari kendala tersebut adalah dengan implementasi sistem rekomendasi produk yang dapat memudahkan pelanggan mendapatkan produk yang mereka butuhkan. Sistem rekomendasi ini menggunakan data histori penjualan manual di salah satu toko Proban Motoparts, dengan total jumlah data transaksi sebanyak 5 transaksi. Implementasi sistem rekomendasi content-based filtering menggunakan algoritma apriori ini untuk menghasilkan produk dengan nilai support tertinggi yang akan direkomendasikan kepada pelanggan. Aplikasi rekomendasi ini memudahkan pelanggan dalam memesan berbagai kebutuhan suku cadang motor yang mereka butuhkan cukup menggunakan perangkat android, hasil persentase pilihan pelanggan diketahui hingga 50.79% memilih sangat setuju dan 35.19% memilih setujuberdasarkan hasil kuesioner dengan 42 responden yang bersedia, dan sebanyak 87.09% merasa puas dengan sistem rekomendasi produk ini.
Deteksi Email Spam menggunakan Algoritma Convolutional Neural Network (CNN) Bachri, Chris Moulana; Gunawan, Wawan
JEPIN (Jurnal Edukasi dan Penelitian Informatika) Vol 10, No 1 (2024): Volume 10 No 1
Publisher : Program Studi Informatika

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26418/jp.v10i1.73306

Abstract

Deteksi email spam merupakan isu penting dalam keamanan siber di Indonesia, yang menempati posisi delapan teratas di dunia dalam hal pengiriman spam. Untuk mengatasi tantangan ini, penelitian ini memperkenalkan penggunaan algoritma Convolutional Neural Network (CNN). Dengan kemampuan superior dalam mempelajari dan mengenali pola dari dataset besar, CNN menawarkan pendekatan berbasis kecerdasan buatan yang lebih efektif daripada metode tradisional. Penelitian ini mengembangkan model CNN dengan menganalisis teks dari 15.271 email berbahasa Inggris dan Indonesia dengan menggunakan teknik pembersihan teks dan Tokenization. Hasilnya menunjukkan keefektivitasan CNN yang signifikan dalam mengklasifikasikan email dengan tingkat akurasi tinggi sebesar 99.67% untuk data uji 20%, 99.64% untuk data uji 30%, dan 99.63% untuk data uji 40%. Berdasarkan hasil pengujian tersebut menunjukkan bahwa algoritma CNN berpotensi kuat dalam meningkatkan keamanan digital.
Pemanfaatan Chi Square dan Ensemble Tree Classifier pada Model SVM, KNN dan C4.5 dalam Penjualan Online Indriyanti, Prastika; Gunawan, Wawan
Faktor Exacta Vol 17, No 3 (2024)
Publisher : LPPM

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30998/faktorexacta.v17i3.24149

Abstract

This research aims to assist MSMEs in overcoming problems in online sales. Currently, sellers only prepare stock without knowing how well the products are sold in their market segment. In the city of Tangerang alone, there are 222,602 MSMEs with various product categories. Therefore, besides utilizing offline sales, business actors should also engage in online sales. This research conducts feature selection using the Chi-Square method and Ensemble Tree Classifier to select the top 6 and 10 features. The SVM, KNN, and C4.5 algorithms are used to build prediction models based on the selected features. Using feature selection, it was found that the influential features are Estimated Shipping Cost, Shipping Cost Paid by Buyer, Total Product Price, and Estimated Shipping Cost Discount. The evaluation results using the three algorithms, SVM, KNN, and C4.5, indicate that the highest accuracy value is obtained when using the C4.5 model with data from the ensemble tree classifier with 6 features at 0.86%, followed by the C4.5 model with 10 features, KNN with 6 features, and KNN with 10 features, all of which source data from the ensemble tree classifier with an accuracy value of 0.85%.
Pemodelan Penentuan Pupuk Menggunakan Metode AHP dan SAW Eliyani, Eliyani; Gunawan, Wawan; Wahyuningram, Nugroho; Triyono, Gandung
Faktor Exacta Vol 17, No 3 (2024)
Publisher : LPPM

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30998/faktorexacta.v17i3.24580

Abstract

The dataset used in this study comprises criteria and fertilizer brands for rice, based on assessments conducted by farmers. The criteria were: price (C1), product (C2), quality (C3), quantity (C4), recommendation (C5), effectiveness (C6), and suitability (C7). Each criterion is weighted as Very Good (4), Good (3), Fair (2), and Poor (1). The evaluated fertilizers consisted of 15 brands: Nitrea, Caping Tani, Pertiphos, NPK padi kuning, SP 36, Meroke, Pusri, Nitroku, Ponska, ZA, Urea, and NPK Pak Tani. The assessments were carried out by distributing questionnaires to 50 farmers who shop at Cv. Sari Alam Tani Store, where farmers could select more than one brand of fertilizer. The most chosen fertilizer by the farmers was Urea. This study aims to verify if Urea is indeed the best fertilizer using the AHP and SAW algorithms based on the established criteria. The results indicate that NPK Padi Kuning was ranked first with a score of 1.72, followed by NPK Tawon and NPK Pak Tani with scores of 1.61 and 1.40, respectively. Urea, despite being the most chosen by farmers, ranks fourth with a score of 1.25.
APPLICATION ARABIC LEARNING BASED ON MULTIMEDIA USING ASYNCHRONOUS METHOD Wawan Gunawan
IJISCS (International Journal of Information System and Computer Science) Vol 4, No 2 (2020): IJISCS (International Journal Information System and Computer Science)
Publisher : Bakti Nusantara Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.56327/ijiscs.v4i2.894

Abstract

Every school has different featured programs but it has similar goals to educate the nation based on faith and piety on Quran and As-Sunnah. Daily conversation will often be used between the school and the student's parents, both basic and later. To obtain maximum results in the learning process, it is necessary to use Arabic communication between the parents and the students outside the school. There are Parents difficulties to speak Arabic daily conversation to support the teaching and learning process at school and home. There is a deadlock between parents and students due to parents' lack of understanding in Arabic. How to helping the realization of the Arabic learning process given to students outside the school environment by repeating the learning process with parents with multimedia application tools and building Arabic learning applications by using asynchronous based multimedia applications.