Ma'ruf, Muhammad Rifqi (2026) Evaluasi Sistem Notifikasi Percakapan Berbasis Konteks dalam Kerangka Kerja AwarNS untuk Manajemen Tugas Mahasiswa. Other thesis, Institut Teknologi Sepuluh Nopember.
|
Text
5025221060-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (9MB) |
Abstract
Mahasiswa menghadapi tuntutan akademik yang kompleks, dinamis, dan sering kali bersamaan dengan distraksi digital, sehingga membutuhkan sistem manajemen tugas yang tidak hanya mengingatkan berdasarkan tenggat waktu, tetapi juga mampu menyesuaikan intervensi dengan kondisi pengguna. Sistem manajemen tugas digital yang umum digunakan masih banyak mengandalkan pengingat dengan tenggat waktu, sehingga notifikasi berpotensi muncul pada konteks yang kurang relevan. Tugas akhir ini merancang dan mengimplementasikan sistem notifikasi percakapan berbasis konteks untuk manajemen tugas mahasiswa menggunakan kerangka kerja AwarNS. Sistem memanfaatkan paradigma Sense-Analyze-Act dengan mengakuisisi konteks pengguna berupa lokasi melalui geofencing, aktivitas fisik, dan durasi penggunaan aplikasi media sosial. Konteks tersebut digunakan untuk menghasilkan intervensi berupa notifikasi teguran. Evaluasi dilakukan menggunakan desain within-subject selama sembilan hari dengan tiga scenario pengujian, yaitu notifikasi berdasarkan tenggat waktu, notifikasi statis berbasis konteks, dan notifikasi generatif berbasis konteks. Metrik evaluasi meliputi task completion rate (TCR), notification dismissal rate (NDR), serta usability berbasis butir pertanyaan dari SUS/CSUQ. Hasil pengujian menunjukkan bahwa scenario notifikasi berdasarkan tenggat waktu memperoleh TCR tertinggi sebesar 81,07%, diikuti skenario notifikasi statis berbasis konteks sebesar 80,00%. Skenario notifikasi statis berbasis konteks juga memperoleh skor pengalaman pengguna tertinggi sebesar 82,17, sedangkan notifikasi generatif menunjukkan tingkat pengabaian ternormalisasi yang tinggi dan berpotensi menambah beban kognitif pengguna. Temuan ini menunjukkan bahwa notifikasi berbasis konteks lebih efektif ketika bersifat reaktif, singkat, dan muncul di momen yang tepat. Sementara itu, intervensi percakapan generatif perlu dirancang secara lebih selektif agar tidak menjadi distraksi baru bagi mahasiswa. =================================================================================================================================
Students face complex, dynamic academic demands, often coupled with digital distractions, creating a need for a task management system that not only reminds them based
on deadlines but is also capable of adjusting interventions to the user's condition. Commonly used digital task management systems still heavily rely on deadline-based reminders, which potentially causes notifications to appear in less relevant contexts. This study designs and implements a context-based conversational notification system for student task management using the AwarNS framework. The system utilizes the Sense-Analyze-Act paradigm by acquiring the user's context in the form of location through geofencing, physical activity, and social media application usage duration. This context is used to generate interventions in the form of reminder notifications. Evaluation was conducted using a within-subject design over nine days with three testing scenarios, namely deadline-based notifications, context-based static notifications, and context-based generative notifications. Evaluation metrics included the task completion rate (TCR), notification dismissal rate (NDR), and usability based on questionnaire constructs from SUS/CSUQ. The test results show that the deadline-based notification scenario obtained the highest TCR of 81,07%, followed by the context-based static notification scenario at 80,00%. The context-based static notification scenario also obtained the highest user experience score of 82,17, while generative notifications showed a high normalized dismissal rate and potentially increased the user's cognitive load. These findings indicate that contextbased notifications are more effective when they are reactive, concise, and appear at the right moment. Meanwhile, generative conversational interventions need to be designed more selectively so as not to become a new distraction for students.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | AwarNS, sistem berbasis konteks, notifikasi percakapan, manajemen tugas, mobile sensing.AwarNS framework, context-aware system, conversational notification, mobile sensing, task management. |
| Subjects: | Q Science > QA Mathematics > QA76.774.A53 Android T Technology > T Technology (General) > T58.8 Productivity. Efficiency |
| Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis |
| Depositing User: | Muhammad Rifqi Ma'ruf |
| Date Deposited: | 04 Aug 2026 02:40 |
| Last Modified: | 04 Aug 2026 02:40 |
| URI: | http://repository.its.ac.id/id/eprint/143195 |
Actions (login required)
![]() |
View Item |
