Claim Missing Document
Check
Articles

Found 2 Documents
Search
Journal : Journal of Computer and Information Systems Ampera

Model Pedukung Keputusan Dengan Metode Analitycal Hierarchy Process: Studi Kasus Proses Bongkar Muatan Barang Kapal Ari Muzakir
Journal of Computer and Information Systems Ampera Vol. 2 No. 2 (2021): Journal of Computer and Information Systems Ampera
Publisher : APTIKOM SUMSEL

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51519/journalcisa.v2i2.75

Abstract

As a center for leaning ships, the Port is a center that plays an important role as a facility that can connect one island to another in trading activities. In this study emphasizes the final goal of (1) determining the factors used as criteria for determining the right and optimal loading decisions, (2) calculating the priority weights of the decision criteria, (3) determining priority weights, (4) obtaining a design that best in priority ship unloading. In this study we use hierarchical based analysis methods or commonly called Analytical hierarchy process (AHP). This method performs pairwise comparisons between criteria and sub criteria by producing a criteria value matrix and produces consistent information each factor from the sub-criteria assessed where the CR <0.1. The results of the scores for each 3 respondents are where the arrival time of the ship scores a total of 17.46 with the highest score, the type of load total score of 11.89, the type of ship the total score is 6.61 and the total loading score of 3.05.
Model Deteksi Berita Palsu Menggunakan Pendekatan Bidirectional Long Short Term Memory (BiLSTM) Ari Muzakir; Uci Suriani
Journal of Computer and Information Systems Ampera Vol. 4 No. 2 (2023): Journal of Computer and Information Systems Ampera
Publisher : APTIKOM SUMSEL

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51519/journalcisa.v4i2.397

Abstract

Isu berita palsu telah menarik perhatian masyarakat dan akademisi. Penyebaran informasi yang tidak akurat berpotensi mengubah pandangan publik dan memungkinkan manipulasi opini. Dalam konteks data yang melimpah, kami mengembangkan model untuk mendeteksi berita palsu dengan mengklasifikasikan fitur linguistik murni. Dengan pendekatan pembelajaran mendalam, kami mengevaluasi respons terhadap artikel tertentu menggunakan model Jaringan Saraf Rekuren Bidirectional Long Short Term Memory (BiLSTM) dan representasi kata dari embeddings GloVe. Hasil evaluasi menunjukkan adaptabilitas model pada data latih dengan kerugian terendah 0.30% dan akurasi tinggi 99.14%. Gabungan antara embeddings GloVe dan Bi-LSTM memunculkan hasil yang positif. Penelitian ini memiliki potensi untuk memberikan kontribusi dalam penanggulangan penyebaran berita palsu yang semakin meresahkan di berbagai bidang.