Muhammad Rifansyah
Unknown Affiliation

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

Perbandingan Tingkat Akurasi Algoritma Naïve Bayes dan Support Vector Machine Dalam Analisis Sentimen Pengguna Aplikasi ShopeePay Pada Google Play Store Hilmi Ammar; Fadli Al Gani; Muhammad Rifansyah; Firman Noor Hasan
Prosiding Seminar Nasional Teknoka Vol 9 (2024): Proceeding of TEKNOKA National Seminar - 9
Publisher : Fakultas Teknik, Universitas Muhammadiyah Prof. Dr. Hamka, Jakarta

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.22236/teknoka.v9i1.17549

Abstract

This research aims to analyze user sentiment towards the ShopeePay application using the Naïve Bayes and SVM algorithms with data obtained through web scraping. Of the 1500 data obtained through scraping, 63 empty data were removed in the cleaning process, leaving 1437 data. This data was then divided into a training set (1149 data) and a test set (288 data). The results showed that the Naïve Bayes algorithm achieved an accuracy of 84.38%, a precision of 79.73%, a recall of 88.72%, and an F1-score of 83.99%, while the Support Vector Machine (SVM) algorithm achieved an accuracy of 80.56%, a precision of 84.07%, a recall of 71.43%, and an F1-score of 77.24%. Overall, Naïve Bayes performed better than Support Vector Machine, especially Naïve Bayes was superior in detecting positive sentiment, while SVM was better in detecting negative sentiment. Data visualization shows that out of 1437 users, around 52.7% gave positive reviews and 47.3% negative reviews, with a diverse rating distribution from users. Based on this distribution, the ShopeePay application user experience can be categorized as predominantly positive, with a difference of 5.4% indicating the difference between 52.7% positive reviews and 47.3% negative reviews from ShopeePay application users.
Transformasi Peran Profesional Pajak di Era Digitalisasi Perpajakan: Systematic Literature Review Muhammad Rifansyah; Christina Fransiska; Muhammad Ichsan Diarsyad; Anggara Aulia Himokta
Media Akuntansi Perpajakan Vol 11, No 1 (2026): Media Akuntansi Perpajakan
Publisher : Universitas 17 Agustus 1945 Jakarta

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52447/map.v11i1.9530

Abstract

This study aims to analyze the transformation of tax professionals’ roles resulting from tax digitalization in Indonesia, particularly the implementation of the Core Tax Administration System. The method used was a Systematic Literature Review employing the SPIDER framework and the PRISMA protocol. A total of 28 articles were selected from 216 identified articles across three academic databases (Google Scholar, Scopus, and SINTA) for the period 2019–2026. The analysis was conducted through thematic synthesis. The findings indicate that digitalization is shifting the role of tax professionals from administrative implementers to data-driven strategic advisors across ten dimensions: primary functions, dominant activities, work patterns, data access, client relationships, relationships with tax authorities, sources of added value, clerical workload, professional risks, and service foundations. Furthermore, the new competencies required include digital hard skills, strategic soft skills, and professional ethics that cannot be replaced by automation. Additionally, the transformation produces a dual impact: increased efficiency and the threat of automation regarding basic compliance tasks. As a result, tax digitalization has transformed the role of tax professionals, not as an existential threat, but as an opportunity to adaptively reposition their roles through digital upskilling and service restructuring, while remaining relevant as a bridge between taxpayers and tax authorities in digitalization era.