Claim Missing Document
Check
Articles

Found 3 Documents
Search

The Role of ChatGPT in Business Information Systems to Support Strategic Decision Making in Medium-Scale Enterprises Karno Diantoro; Era Sari Munthe; Agus Herwanto; Roy Mubarak; Nanik Istianingsih
Jurnal Minfo Polgan Vol. 13 No. 1 (2024): Artikel Penelitian
Publisher : Politeknik Ganesha Medan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33395/jmp.v13i1.13673

Abstract

Medium-scale companies often operate in dynamic and competitive business environments. Faced with market changes and new opportunities, timely and accurate strategic decisions are key to maintaining and enhancing the competitiveness of the company. This research aims to examine the role of ChatGPT in business information systems to support strategic decision-making in medium-scale enterprises. The research method employed is a literature review with a qualitative approach and descriptive analysis. Descriptive analysis will be used to present information from systematically selected articles from Google Scholar within the timeframe of 2014-2024. The study results indicate that in the continuously evolving digital era, medium-scale companies increasingly rely on technology to address increasingly complex business challenges. One innovation playing a significant role in today's business landscape is the presence of ChatGPT. As an artificial intelligence model capable of understanding and generating text naturally, ChatGPT has a significant impact on business information systems and supports strategic decision-making in these companies.
Sistem Pakar untuk Identifikasi Risiko Proyek Teknologi Informasi Berbasis Metode Fuzzy Logic Era Sari Munthe; Anwar T. Sitorus; Franky Gerald Cliford Manoppo; Devi Puspita Sari; Filda Angellia
Jurnal Minfo Polgan Vol. 13 No. 2 (2024): Artikel Penelitian
Publisher : Politeknik Ganesha Medan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33395/jmp.v13i2.14217

Abstract

Dalam pengembangan proyek teknologi informasi, identifikasi risiko menjadi tahap penting yang dapat memengaruhi keberhasilan proyek. Risiko yang tidak diidentifikasi sejak awal dapat menyebabkan peningkatan biaya, keterlambatan waktu, atau bahkan kegagalan total proyek. Penelitian ini bertujuan untuk mengembangkan sistem pakar berbasis metode Fuzzy Logic untuk identifikasi risiko proyek teknologi informasi secara akurat dan dinamis. Sistem ini dirancang untuk membantu manajer proyek dan tim dalam mengantisipasi serta mengelola potensi risiko dengan lebih efektif. Metode Fuzzy Logic dipilih karena kemampuannya dalam mengolah data yang bersifat tidak pasti dan ambigu, yang sering kali muncul dalam penilaian risiko. Penggunaan Fuzzy Logic memungkinkan penilaian risiko yang lebih fleksibel, dengan mempertimbangkan berbagai faktor risiko proyek seperti kompleksitas teknis, ketidakpastian anggaran, dan jadwal proyek. Sistem pakar ini terdiri dari beberapa tahapan utama: pengumpulan data risiko, penyusunan aturan fuzzy, dan penerapan metode inferensi fuzzy untuk menentukan tingkat risiko. Data risiko diperoleh melalui wawancara dengan pakar proyek teknologi informasi serta tinjauan literatur terkait. Hasil uji coba menunjukkan bahwa sistem pakar berbasis Fuzzy Logic ini mampu mengidentifikasi tingkat risiko proyek dengan akurasi tinggi dan menyediakan informasi yang bermanfaat untuk pengambilan keputusan. Diharapkan bahwa implementasi sistem ini dapat meminimalkan risiko yang tidak diantisipasi dan mendukung keberhasilan proyek teknologi informasi di berbagai sektor. Dengan demikian, sistem pakar ini berpotensi menjadi alat bantu yang efektif bagi manajer proyek dalam mengelola risiko proyek secara proaktif.
LLM-Assisted Versus Developer-Written Unit Tests in Java Applications: Comparing Mutation Scores, Branch Coverage, and Test Development Time Anisah Mutiara Zahra Hanifah; Era Sari Munthe
Digitus : Journal of Computer Science Applications Vol. 4 No. 2 (2026): April 2026
Publisher : PT Penerbit Ilmiah Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.61978/digitus.v4i2.2024

Abstract

Large language models (LLMs) can assist unit test development, but faster production of candidate tests may not translate into more effective usable test suites. This study analyzes a balanced repeated-measures dataset comparing LLM-assisted and developer-written Java unit tests across mutation effectiveness, structural coverage, and development effort. The supplied records represent 32 developer profiles, 24 modules grouped into six projects, and 192 tasks, with 96 tasks per condition. The outlined study period was November 2025–February 2026 in Semarang, Indonesia; recorded task dates span December 2025–January 2026. The primary outcome assigns zero effectiveness to unusable suites while retaining strict mutation scores for usable suites. Mean effective mutation scores were 55.22% for LLM-assisted tasks and 52.54% for manual tasks. The workbook's adjusted difference was 2.68 percentage points (95% CI −2.43 to 7.79; p = .304). Usable suites were less frequent with LLM assistance (87.5% versus 94.8%), whereas conditional branch coverage was higher (81.54% versus 74.40%). Mean effort was 26.36 versus 33.00 minutes, with an adjusted geometric mean ratio of 0.787 (95% CI 0.747–0.829). These records illustrate how an apparent advantage among usable suites can diminish when delivery failures are included. Contradictory provenance labels prevent interpreting the supplied records as verified empirical observations pending source reconciliation.