The rapid development of digital technology has driven various innovations in the financial sector, one of which is cryptocurrency, which is increasingly discussed on social media and various digital platforms. On the other hand, price volatility, security issues, and evolving regulations have led to diverse public opinions regarding cryptocurrency. Therefore, this study aims to analyze public …
The internship application process at the Management Informatics Study Program of Sriwijaya University still uses Google Form with document management that is not yet integrated, causing the administration and monitoring process to become less effective. This study aims to develop a web-based Internship Application Information System (SIKP) that can support the internship submission, verificati…
The management of radiology examination data at Klinik Su’Adah is still carried out manually using notebooks and Microsoft Excel files, causing difficulties in patient data retrieval, examination result storage, and report generation. This condition also increases the risk of recording errors and data loss. This study aims to design and develop a web-based radiology data management applicatio…
Brain tumor segmentation from Magnetic Resonance Imaging (MRI) is crucial for accurate diagnosis and therapy planning. CNN architectures like U-Net are considered the gold standard, but they have intrinsic limitations in processing long-range spatial context globally. This study proposes the use of a Vision Transformer architecture, specifically UNETR, and compares its performance against a bas…
The rapid growth of digital wallets in Indonesia, particularly GoPay, has generated tens of thousands of user reviews containing valuable information about service satisfaction. This study aims to classify GoPay application user satisfaction into positive, neutral, and negative categories using the Fine-Tuning IndoBERT (Indonesian BERT) method. As a Transformer-based model trained specifically …
ChatGPT is an artificial intelligence technology that triggers diverse public opinions on social media, particularly on the X platform. The high volume of data and unstructured text characteristics make manual mapping of public perception difficult. Therefore, this study aims to develop an automated sentiment analysis system to classify public opinion toward ChatGPT into Negative, Neutral, and …
The phenomenon of Papi Abe as a family content creator who has gone viral on TikTok has generated a wide range of public responses, both positive and negative, particularly regarding father-child interactions in parenting content and the potential for child exploitation. Therefore, it is important to understand public opinions toward Papi Abe as a basis for evaluating and improving content qual…
This research aims to analyze public sentiment toward the early leadership of President Prabowo Subianto using the IndoBERT language model. Public opinions were collected from various social media platforms, including X, YouTube and TikTok, resulting in a dataset of 78,851 texts labeled into positive and negative sentiments. The data underwent preprocessing stages such as cleaning, case folding…
Healthy and diseased mango plants are difficult to distinguish at early stages due to insignificant differences in leaf color, texture, and spot patterns, which often leads to delayed identification and results in decreased mango production and fruit quality. This study develops an Android-based application to detect diseases in mango plants based on leaf images using the YOLO11 algorithm. The …
Prediction is the process of estimating future conditions using historical data as the basis for decision-making. This research develops a web-based fish catch prediction system using XGBoost algorithm based on time series with nonlinear regression method to predict fish production results submitted to the port. The data used is aggregate fish catch data from 12 fishing points during 2020– 20…
Formula 1 qualifying sessions play a crucial role in determining race outcomes, as grid position strongly influences competitive advantage, particularly on circuits with limited overtaking opportunities. This study aims to implement the Light Gradient Boosting Machine (LightGBM) algorithm to predict drivers’ qualifying lap times in Formula 1 using historical time-series data from the 2018–2…
This study implements a Conditional Generative Adversarial Network (CGAN) to enhance the contrast of fetal abdominal ultrasound images. The novelty lies in increasing the Reconstruction Loss weight (λL1) to 1000 to ensure structural stability in medical data. Using U-Net and PatchGAN architectures, the model was trained against CLAHE targets. The results demonstrate significant enhancement wit…
Nutrition facts tables on food packaging are a crucial source of information for consumers, but manual data extraction is often inefficient and prone to errors due to varied table structures and small text sizes. This research proposes an automated system for text detection and recognition in nutrition tables by integrating two advanced deep learning models: PaddleOCR and TrOCR. PaddleOCR is em…
Adaptive Artificial Intelligence (AI) plays a significant role in shaping player experience in digital games, particularly in action-based combat scenarios where behavioral dynamics influence perceived challenge, fairness, and engagement. This study investigates the implementation of a Fuzzy State Machine (FuSM) compared to a conventional Finite State Machine (FSM) for boss behavior in a 2D mul…
his study aims to analyze user sentiment toward the Gojek application using the Long Short-Term Memory (LSTM) method. Review data was obtained through scraping the Gojek application service, resulting in a dataset that was then classified into positive and negative sentiment categories. The data underwent class balancing using the undersampling technique, followed by a pre-processing stage that…
The large and unstructured volume of IMDb movie review data makes it difficult to map the opinions of the audience manually. Therefore, this study aims to develop a sentiment analysis system using BERT to classify IMDb movie reviews into positive and negative categories. The data used is secondary data amounting to 50,000 English-language data. This research stage includes pre-processing of tex…
Hepatitis is one of the global health problems with a persistently high incidence rate and the potential to cause serious complications if not detected early. Meanwhile, the use of laboratory data for machine learning-based diagnosis still faces challenges such as missing values, imbalanced class distribution, and the limited application of optimal combinations of feature selection methods and …
The development of blockchain technology has increased cryptocurrency transaction activities, especially Ethereum, whose daily transaction volume is highly fluctuating. This condition makes transaction volume prediction important for understanding network activity patterns. This study aims to develop a prediction model for daily Ethereum transaction volume using the XGBoost Regressor algorithm.…
Identifying bacteria traditionally takes 24–72 hours and is prone to human error. With bacterial infections and antimicrobial resistance causing over 1 million deaths annually, faster and more accurate methods are urgently needed. This study compares two deep learning models, ResNet-50 (a classic CNN) and ConvNeXt-Tiny (a modern CNN) for classifying microscopic bacteria images using the DIBaS…
The increasing demand for laptops in the digital era is often not aligned with the ease of choosing the right specifications due to the vast variety of brands and features available in the market. This research aims to develop an objective Decision Support System (DSS) for laptop selection by integrating the Rank Order Centroid (ROC) method for criteria weighting and Multi-Objective Optimizatio…
The development of wearable technology has increased the use of smartwatches as devices that support daily activities and health monitoring. However, the large number of product choices with conflicting specifications and technical criteria makes the smartwatch selection process complex. This study aims to develop a web-based Decision Support System (DSS) for smartwatch selection by integrating…
Digital game development requires efficient pathfinding algorithms. This study compares A* and Jump Point Search (JPS) in a hill-climbing adventure game based on Roblox. The three-dimensional terrain was converted into a 181x209 uniform-cost two-dimensional grid with a cell size of 4 Roblox units. Both algorithms were implemented in Roblox Studio using Lua and the octile distance heuristic for …
This research aims to develop an automated sentiment analysis model using a hybrid method of Random Forest (RF) and Support Vector Machine (SVM) to evaluate DANA application user reviews on the Google Play Store. A total of 50,001 secondary review data from Kaggle were used in this study, partitioned into 80% training data and 20% testing data. The analysis process includes text preprocessing a…
Conventional talent scouting systems for potential swimming athletes still rely on subjective assessments by coaches, potentially leading to bias and measurement errors. To address this issue, this study developed a Hybrid Backpropagation–Genetic Algorithm-based classification system to identify potential athletes based on distance specialization. The data used consisted of 100 potential athl…
Accurate mental state identification is important across various fields, and EEG is a biomedical signal widely used for this purpose as it can directly and non-invasively record brain electrical activity. However, the complex, nonlinear, and high-dimensional characteristics of EEG signals require effective classification methods. This study aims to implement and compare the performance of Logis…
This study develops NusantaraLaw, a Domain-Adaptive Retrieval-Augmented Generation (RAG) system for Indonesian law via QLoRA fine-tuning of the Qwen 3.5 9B model on a national legal corpus. A dataset of 6,502 instruction-response pairs was curated from eight official legal sources and validated by two legal experts via Stratified Random Sampling (250 samples, margin of error 6.96 percent). Four…
This study investigates automated road damage detection by combining the YOLOv11 algorithm with the Slicing Aided Hyper Inference (SAHI) approach on the multinational RDD2022 dataset, covering four damage categories: Longitudinal Crack (D00), Transverse Crack (D10), Alligator Crack (D20), and Pothole (D40). Ten YOLOv11 training experiments and eleven SAHI hyperparameter configurations were cond…
This study aims to design and develop a web-based Posyandu Service Information System and Health Monitoring System for mothers and toddlers in Jaga Baya Village. The study was conducted because Posyandu data management was still carried out manually through record-keeping in books. The research methods used in this study included observation, interviews, and literature review, while the system …
Mobile network operator services play a crucial role in fulfilling the internet needs of the public. However, the large number of available options and the variety of packages often make it difficult for users to choose a suitable operator. This issue is experienced by students as newcomers in Timbangan Village, North Indralaya District, due to limited information regarding the quality of opera…