ITB Journal
Not a member yet
5999 research outputs found
Sort by
Foundations of Domain-specific Large Language Models for Islamic Studies: A Comprehensive Review
Large language models (LLMs) have undergone rapid evolution and are highly effective in tasks such as text generation, question answering, and context-driven analysis. However, the unique requirements of Islamic studies, where textual authenticity, diverse jurisprudential interpretations, and deep semantic nuances are critical, present challenges for general LLMs. This article reviews the evolution of neural language models by comparing the historical progression of general LLMs with emerging Islamic-specific LLMs. We discuss the technical foundations of modern Transformer architectures and examine how recent advancements, such as GPT-4, DeepSeek, and Mistral, have expanded LLM capabilities. The paper also highlights the limitations of standard evaluation metrics like perplexity and BLEU in capturing doctrinal, ethical, and interpretative accuracy. To address these gaps, we propose specialized evaluation metrics to assess doctrinal correctness, internal consistency, and overall reliability. Finally, we outline a research roadmap aimed at developing robust, ethically aligned, and jurisprudentially precise Islamic LLMs
Menavigasi Transisi ITB Innovation Park melalui Perspektif Multi Tingkat
The development of the science techno park (STP) in Indonesia aims to facilitate innovation-based industrial growth. Concurrently, the Bandung Institute of Technology (ITB) is committed to realizing an entrepreneurial university by leveraging research results through the ITB-STP, also known as the ITB Innovation Park (IIP). The IIP acts as an innovation center that connects research with market needs and functions as an innovation commercialization accelerator. Research using the Multi-Level Perspective Framework indicates policy shifts in STP development during the 2015– 2019 and 2020–2024 periods. Despite producing various innovative products, IIP faces challenges in transitioning the socio-technical system for commercializing research outputs, particularly in integrating with the broader regime and ensuring innovation sustainability. IIP’s initiatives demonstrate instances of fit & fit; however, during the fit & stretch phase, IIP continues to encounter challenges, particularly concerning market responses to the innovative products offered. This indicates that the stretching process remains in progress. Consequently, IIP’s success in reshaping the regime depends on enhanced collaboration, systematic innovation validation, and supportive government policies to foster the establishment of a new innovation-based regime.Pengembangan Science Techno Park (STP) di Indonesia dilakukan sebagai upaya untuk memfasilitasi pertumbuhan industri berbasis inovasi. Di sisi lain, ITB berkomitmen untuk mewujudkan entrepreneurial university dengan mengoptimalkan potensi inovasi dari hasil riset melalui STP ITB atau ITB Innovation Park (IIP). Innovation Park (IIP) berperan sebagai pusat inovasi yang menghubungkan riset dengan kebutuhan pasar dan berfungsi sebagai akselerator komersialisasi inovasi. Hasil penelitian dengan menggunakan kerangka Multi-Level Perspective menunjukkan pergeseran dalam perubahan kebijakan terkait konsep pengembangan STP dalam periode 2015- 2019 dan 2020-2024. Meskipun Innovation Park (IIP) telah menghasilkan berbagai produk inovasi, transisi sistem sosio-teknikal dalam komersialisasi produk riset masih menghadapi tantangan. Upaya IIP menunjukkan terjadinya fit & fit, namun dalam proses fit & stretch IIP masih menghadapi tantangan, salah satunya terkait respons pasar terhadap produk-produk inovasi yang ditawarkan sekaligus mencerminkan proses stretch yang masih berlangsung. Oleh karena itu, keberhasilan IIP untuk mengubah rezim membutuhkan peningkatan kolaborasi, validasi inovasi, serta dukungan kebijakan pemerintah dalam membentuk rezim baru yang berbasis inovasi
Rancangan Dinding Interaktif Vokasi Berbasis Human Centered Design dengan Pengukuran User Experience Questionnaire
The Augmented Reality (AR)-based Vocational Interactive Wall is designed to showcase content from the Directorate General of Vocational Education, including profiles and innovative works through an interactive wall. This application aims to provide a more immersive and engaging experience, allowing visitors to play, learn, and have fun simultaneously using AR technology. The purpose of this research is to assist students and visitors in exploring the profile of the Directorate General of Vocational Education through an AR-based interactive medium. The study employs the Human-Centered Design (HCD) method for product development and the User Experience Questionnaire (UEQ) for evaluating user satisfaction. Data were collected from 20 respondents during the Vokasi Fest x Kampus Merdeka 2023 event. The analysis results indicate that the AR-based Vocational Interactive Wall application is deemed "suitable" for use, offering an engaging and interactive user experience. These findings highlight the application's potential as an innovative learning medium that effectively enhances engagement in educational settings.Dinding Interaktif Vokasi berbasis Augmented Reality (AR) dirancang untuk menampilkan konten dari Direktorat Jenderal Pendidikan Vokasi, meliputi profil dan karya inovatif melalui dinding interaktif. Aplikasi ini bertujuan menjadi pengalaman yang lebih imersif dan menarik, memungkinkan pengunjung untuk bermain, belajar, dan bersenang-senang secara bersamaan dengan teknologi AR. Penelitian ini bertujuan membantu pelajar dan pengunjung mengeksplorasi profil Direktorat Jenderal Pendidikan Vokasi melalui media interaktif berbasis AR. Metode yang digunakan adalah Human-Centered Design (HCD) untuk pengembangan produk dan User Experience Questionnaire (UEQ) untuk evaluasi kepuasan pengguna. Data diperoleh dari 20 responden di acara Vokasi Fest x Kampus Merdeka 2023. Hasil analisis menunjukkan aplikasi Dinding Interaktif Vokasi berbasis AR dinilai “Layak” untuk digunakan, dengan pengalaman pengguna yang interaktif dan menarik. Temuan ini menunjukkan potensi aplikasi sebagai media pembelajaran inovatif yang efektif dalam meningkatkan keterlibatan di lingkungan pendidikan
Characteristics and Patterns of Urbanization of Non-Urban Regions: the case of Kudus Regency in Central Java Province, Indonesia
Urbanization in Java is characterized by processes that occur outside the boundaries of cities, particularly in the territories of regencies that are still perceived as non-urban districts. Using the case of Kudus Regency in Central Java Province, this study analyzed the urbanization process that takes place in Javanese regencies, highlighting the importance of the process outside of cities, especially in regencies with high urbanization levels. The analysis showed that the urbanization process is reflected in the spatial expansion of urbanized villages and the growing urban population in growing urban areas in the regency. The pattern and characteristics of urbanization in Kudus Regency are relatively independent from the influence of major cities that are located more than 60 km away, and the process has been observed as accommodating the growing population and its activities in growing urban areas. The spatial growth in the urbanization process corresponds with the growth in population density, supported by the availability of three spatial elements: a high-density core, lesser-density areas close to the core, and road networks connecting them. The analysis further showed that the urbanization process that takes place in Kudus Regency is an in-situ urbanization process in the capital that leads to the formation of a larger city, with the economic and demographic factors of the region supporting and determining the speed and direction of the process. The fast rate of the urbanization process and the expansion of urban areas in the regency are challenging, necessitating management in the same manners as in cities
Optimal Control for Resource Allocation in a Multi-Patch Epidemic Model with Human Dispersal Behavior
A multi-patch epidemic compartmental model with human dispersal behavior studies the spread of the disease and it sets the model to real-world situations. The mobility matrix (M) applies human dispersal behavior to the model. The optimal control theory assists in controlling the disease burden while minimizing the cost of infected individuals and implementing control measures. We formulate a multi-patch SIR model with human dispersal behavior to control and reduce communicable disease outbreaks such as COVID-19 by optimizing resource allocation in Sri Lanka. Results are represented by using the reproduction number (R0), effective reproduction number (Rt), and final epidemic size (ci). Compared to the basic reproduction number (R0), the effective reproduction number (Rt) represents the significant result in the epidemiological model incorporated with control measures. The average number of secondary cases concerning the current susceptible population is represented by Rt and the final epidemic size represents the patched-specified cost value for infected individuals. According to the results, the disease burden can be controlled by vaccination relative to social distancing
Pengembangan Fitur Tampilan Status Dosen Berbasis Raspberry Pi dengan Metode Rapid Application Development (RAD) di Lingkungan Akademik
The advancement of information technology drives improvements in academic services, including providing fast and transparent access to lecturer attendance status information. This study develops a Raspberry Pi-based lecturer status display to show real-time lecturer attendance. The system is developed using the Rapid Application Development (RAD) method through the stages of data collection, analysis, design, implementation, and testing. Functional testing results show a 98% accuracy rate in data synchronization between the database and the display screen. A questionnaire distributed to 12 lecturers revealed that 86.7% of respondents found the application highly beneficial in quickly and transparently obtaining lecturer attendance information. Thus, this application can enhance efficiency and information transparency in the academic environmentKemajuan teknologi informasi mendorong peningkatan layanan akademik, termasuk penyediaan akses informasi yang cepat dan transparan terkait status kehadiran dosen. Penelitian ini mengembangkan tampilan status dosen berbasis Raspberry Pi untuk menampilkan kehadiran dosen secara real-time. Sistem dikembangkan dengan metode Rapid Application Development (RAD) melalui tahapan pengumpulan data, analisis, perancangan, implementasi, dan pengujian. Hasil pengujian fungsionalitas menunjukkan tingkat akurasi 98% dalam sinkronisasi data antara database dan tampilan layar. Kuesioner yang diberikan kepada 12 dosen menunjukkan 86,7% responden merasa aplikasi ini sangat membantu dalam memperoleh informasi kehadiran dosen secara cepat dan transparan. Dengan demikian, aplikasi ini dapat meningkatkan efisiensi dan keterbukaan informasi di lingkungan akademik
Implementasi Convolutional Neural Network Berbasis Model untuk Klasifikasi Kelayakan Citra Permen Jahe pada Perangkat Android
Ginger is one of the primary ingredients for ginger candy. The manual process of evaluating the feasibility of ginger candy at the Tasacika Company is still prone to errors and is less efficient. This research aims to develop a Convolutional Neural Network model for classifying the feasibility of ginger candy and create an Android-based application that facilitates this process. The research method uses an Experimental approach. Model development is carried out with a Convolutional Neural Network with the MobileNetV2 architecture, using the Cross Industry Standard Process for Data Mining methods. Software development is done using the Prototyping method. This research used a dataset of images taken directly from the Tasacika Company's ginger candy factory. The model is trained and tested using Google Colab with the Python programming language and the TensorFlow and Keras libraries. Implementation is carried out using Kotlin and XML. It can be concluded that the research has succeeded in developing a ginger candy feasibility classification model. The test results show that the developed model is effective in minimizing human error in the process of checking the feasibility of ginger candy. This research also succeeded in developing an Android-based ginger candy feasibility classification applicationJahe merupakan salah satu bahan dasar pembuatan permen jahe. Proses evaluasi kelayakan permen jahe di Perusahaan Tasacika yang masih dilakukan secara manual, masih rawan terjadi kesalahan dan kurang efisien. Penelitian ini bertujuan untuk mengembangkan model Convolutional Neural Network untuk klasifikasi kelayakan permen jahe dan membuat aplikasi berbasis android yang memudahkan proses tersebut. Metode penelitian menggunakan pendekatan eksperimental. Pengembangan model dilakukan dengan Convolutional Neural Network dengan arsitektur MobileNetV2, dan menggunakan metode Cross Industry Standard Process for Data Mining. Pengembangan perangkat lunak dilakukan dengan metode Prototyping. Penelitian ini menggunakan dataset berupa gambar yang diambil langsung dari pabrik permen jahe milik Perusahaan Tasacika. Model dilatih dan diuji menggunakan Google Colab dengan bahasa pemrograman Python, serta library TensorFlow dan Keras. Implementasi dilakukan menggunakan Kotlin dan XML. Dapat disimpulkan bahwa penelitian ini telah berhasil mengembangkan model klasifikasi kelayakan permen jahe. Hasil pengujian menunjukkan bahwa model yang dikembangkan efektif dalam meminimalisir human error pada proses pengecekan kelayakan permen jahe. Penelitian ini juga berhasil mengembangkan aplikasi klasifikasi kelayakan permen jahe berbasis Android
Analisis Desain Pekerjaan Peninggian Jalan pada Lokasi Banjir (Studi Kasus: Ruas Bts. Kab. Serang – Bts. Kota Pandeglang)
Abstrak
Ruas Jalan Bts. Kab. Serang – Bts. Kota Pandeglang STA. 1+100 merupakan ruas jalan yang rawan banjir di Provinsi Banten. Alternatif penanganan ruas jalan yaitu berupa peninggian jalan. Penelitian dilakukan dengan perhitungan menggunakan Pedoman Desain Drainase Jalan Nomor: 23/SE/Db/2021 terkait tinggi luapan air banjir sebagai dasar penentuan tinggi minimum peninggian. Analisis desain sistem perkerasan metode empiris menggunakan Manual Desain Perkerasan Jalan Tahun 2024 No. 03/M/BM/2024, lalu dimodelkan untuk dianalisis dengan metode mekanistik-empiris program kenpave dan dilakukan perbandingan dalam aspek biaya dari setiap sistem perkerasan. Hasil analisis kebutuhan peninggian jalan yaitu minimum sebesar 23 cm. Analisis metode empiris didapatkan tebal beton pada rigid pavement yaitu 315 mm dipilih Tipe 1R dan 2R, tebal aspal pada flexible pavement yaitu 285 mm. Ditentukan 2 tipe rigid pavement dan 4 tipe flexible pavement dianalisis dengan program Kenpave. Analisis metode mekanistik-empiris rigid pavement memilih Tipe 2R dengan index cracking dan max stress terkecil, untuk flexible pavement dipilih Tipe 4F dengan jumlah repetisi beban pada fatigue cracking dan permanent deformation yang paling besar. Analisis biaya untuk umur rencana 40 tahun rigid pavement Tipe 2R senilai Rp. 10.615.725,00/m’ dan flexible pavement Tipe 4F senilai Rp. 12.437.698,00/m’. Dipilih Tipe 2R dengan biaya lebih kecil dari Tipe 4F.
Kata kunci : Peninggian jalan, sistem perkerasan, MDP 2024, kenpave.
Abstract
The road section Bts. Kab. Serang – Bts. Kota Pandeglang STA. 1+100 is a flood-prone road in Banten Province. An alternative solution for the road section is to raise the road. The research was conducted with calculations using the Road Drainage Design Guidelines Number: 23/SE/Db/2021 related to the flood overflow height, which serves as the basis for determining the minimum elevation height. The analysis of the pavement design system using the empirical method follows the Road Pavement Design Manual 2024 No. 03/M/BM/2024, which is then modeled for analysis using the mechanistic-empirical method with the Kenpave program, and a cost comparison is conducted for each pavement system. The analysis results for the road elevation requirement indicate a minimum of 23 cm. The empirical method analysis obtained a concrete thickness of 315 mm for rigid pavement, selecting Type 1R and 2R, and an asphalt thickness of 285 mm for flexible pavement. Two types of rigid pavement and four types of flexible pavement were analyzed using the Kenpave program. The analysis of the mechanistic-empirical method for rigid pavement selects Type 2R with the smallest cracking index and max stress, while for flexible pavement, Type 4F is chosen with the highest number of load repetitions in fatigue cracking and permanent deformation. Cost analysis for a 40-year design life of rigid pavement Type 2R is Rp. 10,615,725.00/m’ and flexible pavement Type 4F is Rp. 12,437,698.00/m’. Type 2R was chosen because its cost is lower than Type 4F.
Keywords : Raising road, pavement systems, MDP 2024, kenpave.Abstrak
Ruas Jalan Bts. Kab. Serang – Bts. Kota Pandeglang STA. 1+100 merupakan ruas jalan yang rawan banjir di Provinsi Banten. Alternatif penanganan ruas jalan yaitu berupa peninggian jalan. Penelitian dilakukan dengan perhitungan menggunakan Pedoman Desain Drainase Jalan Nomor: 23/SE/Db/2021 terkait tinggi luapan air banjir sebagai dasar penentuan tinggi minimum peninggian. Analisis desain sistem perkerasan metode empiris menggunakan Manual Desain Perkerasan Jalan Tahun 2024 No. 03/M/BM/2024, lalu dimodelkan untuk dianalisis dengan metode mekanistik-empiris program kenpave dan dilakukan perbandingan dalam aspek biaya dari setiap sistem perkerasan. Hasil analisis kebutuhan peninggian jalan yaitu minimum sebesar 23 cm. Analisis metode empiris didapatkan tebal beton pada rigid pavement yaitu 315 mm dipilih Tipe 1R dan 2R, tebal aspal pada flexible pavement yaitu 285 mm. Ditentukan 2 tipe rigid pavement dan 4 tipe flexible pavement dianalisis dengan program Kenpave. Analisis metode mekanistik-empiris rigid pavement memilih Tipe 2R dengan index cracking dan max stress terkecil, untuk flexible pavement dipilih Tipe 4F dengan jumlah repetisi beban pada fatigue cracking dan permanent deformation yang paling besar. Analisis biaya untuk umur rencana 40 tahun rigid pavement Tipe 2R senilai Rp. 10.615.725,00/m’ dan flexible pavement Tipe 4F senilai Rp. 12.437.698,00/m’. Dipilih Tipe 2R dengan biaya lebih kecil dari Tipe 4F.
Kata kunci : Peninggian jalan, sistem perkerasan, MDP 2024, kenpave.
Abstract
The road section Bts. Kab. Serang – Bts. Kota Pandeglang STA. 1+100 is a flood-prone road in Banten Province. An alternative solution for the road section is to raise the road. The research was conducted with calculations using the Road Drainage Design Guidelines Number: 23/SE/Db/2021 related to the flood overflow height, which serves as the basis for determining the minimum elevation height. The analysis of the pavement design system using the empirical method follows the Road Pavement Design Manual 2024 No. 03/M/BM/2024, which is then modeled for analysis using the mechanistic-empirical method with the Kenpave program, and a cost comparison is conducted for each pavement system. The analysis results for the road elevation requirement indicate a minimum of 23 cm. The empirical method analysis obtained a concrete thickness of 315 mm for rigid pavement, selecting Type 1R and 2R, and an asphalt thickness of 285 mm for flexible pavement. Two types of rigid pavement and four types of flexible pavement were analyzed using the Kenpave program. The analysis of the mechanistic-empirical method for rigid pavement selects Type 2R with the smallest cracking index and max stress, while for flexible pavement, Type 4F is chosen with the highest number of load repetitions in fatigue cracking and permanent deformation. Cost analysis for a 40-year design life of rigid pavement Type 2R is Rp. 10,615,725.00/m’ and flexible pavement Type 4F is Rp. 12,437,698.00/m’. Type 2R was chosen because its cost is lower than Type 4F.
Keywords : Raising road, pavement systems, MDP 2024, kenpave
Formulasi dan Karakterisasi Minyak Atsiri Pala (Myristica fragrans) di Dalam Sistem Nanostructured Lipid Carrier
Essential oil extracted from nutmeg (Myristica fragrans) contains various beneficial compounds that can be used in medicines and cosmetics. Nevertheless, nutmeg essential oil (NEO) use in pharmaceuticals is limited by its high lipophilicity and tendency to degrade. Encapsulating NEO into a nanostructured lipid carrier (NLC) can be an alternative to overcome these problems. Thus, this study aimed to develop an NEO formula in an NLC system (NEO-NLC) for topical use, to characterize NEO-NLC, and to elucidate its potential as an antioxidant. The optimization of the NEO-NLC formula included selecting liquid lipid components, adjusting their ratios, choosing a sonicator, selecting a surfactant system, and optimizing sonication amplitude. The optimum NEO-NLC formula was then characterized by determining particle size and distribution, measuring zeta potential, estimating encapsulation efficiency, and evaluating antioxidant activity. The optimum NEO-NLC formula was obtained with a composition of 1.8% cetyl alcohol, 3.0% almond oil, 1.2% NEO, 3.2% Tween 80, 0.8% Plantacare® 1200, and demineralized water. The resulting NEO-NLC exhibited a particle size of 124.4 ± 4.7 nm, a polydispersity index of 0.35 ± 0.05, a zeta potential of -39.0 ± 3.6 mV, and an encapsulation efficiency of 95.3 ± 0.6%. Encapsulation of NEO into an NLC system resulted in a 2.9-fold increase in antioxidant activity compared to unencapsulated NEO. The developed NEO-NLC showed potential for pharmaceutical or cosmetic raw material use.Minyak atsiri yang diekstraksi dari biji pala (Myristica fragrans) mengandung berbagai senyawa bermanfaat yang dapat dimanfaatkan sebagai bahan baku obat maupun kosmetik. Namun, penggunaan minyak atsiri pala (NEO) di bidang farmasi dibatasi oleh sifat lipofilisitasnya yang tinggi dan kecenderungannya untuk terdegradasi. Enkapsulasi NEO ke dalam suatu nanostructured lipid carrier (NLC) dapat menjadi alternatif untuk mengatasi permasalahan tersebut. Dengan demikian, penelitian ini bertujuan untuk mengembangkan formula NEO dalam suatu sistem NLC (NEO-NLC) untuk penggunaan secara topikal serta melakukan karakterisasi NEO-NLC termasuk elusidasi potensinya sebagai antioksidan. Optimasi formula NEO-NLC dilakukan melalui pemilihan komponen lipid cair, optimasi rasio komponen lipid, pemilihan tipe sonikator, pemilihan sistem surfaktan, dan optimasi amplitudo sonikasi. Formula NEO-NLC optimum kemudian dikarakterisasi meliputi penentuan ukuran partikel dan distribusinya, pengukuran potensial zeta, dan penentuan efisiensi enkapsulasi, serta elusidasi aktivitas antioksidan. Formula NEO-NLC optimum diperoleh dengan komposisi 1,8% setil alkohol, 3,0% minyak almon, 1,2% MAP, 3,2% Tween 80, 0,8% Plantacare® 1200, dan air demineralisasi. NEO-NLC yang dihasilkan menunjukan ukuran partikel sebesar 124,4 ± 4,7 nm, indeks polidispersitas 0,35 ± 0,05, potensial zeta -39,0 ± 3,6 mV, dan efisiensi enkapsulasi 95,3 ± 0,6%. Enkapsulasi NEO ke dalam sistem NLC menghasilkan peningkatan aktivitas antioksidan sebesar 2,9 kali dibandingkan dengan NEO yang tidak dienkapsulasi. NEO-NLC yang dikembangkan menunjukan potensi untuk digunakan sebagai bahan baku farmasi maupun kosmetik