Publication Search

Traditional association rule analysis is effective at uncovering co-purchase patterns but fails to provide a global structural view of the market, which often results in fragmented and isolated insights. This study proposes a hybrid framework that integrates the Apriori algorithm with a Minimum Spanning Tree (MST) in order to validate and contextualize association rules within a single structural backbone. Transaction data from a retail store are transformed into a weighted, undirected product graph using an inverse-support function, and an MST is then extracted to represent the market backbone, while frequent itemsets and strong rules are obtained using Apriori. Experimental results on 236 multi-item transactions show that the MST backbone comprises 10 products and 9 fundamental links, with 66.67% of these links being confirmed by strong association rules, indicating a substantial coherence between statistical and structural evidence. The proposed model identifies 41 Apriori patterns that can be embedded in the MST and ranks them using a new metric, Structural Distance, which enables the categorization of Core Patterns, Bridge Patterns, and Complex Patterns according to their structural tightness. This hybrid perspective distinguishes dense, strategically meaningful bundles from anomalous but frequent combinations that are structurally peripheral, thereby offering a more holistic and actionable alternative to conventional Market Basket Analysis. The validated framework can support various applications, including store layout optimization, cross-selling strategies, and the design of path-based recommender systems, and it opens avenues for future extensions based on dynamic graphs and Graph Neural Networks.

Manual quality assessment of Platelet Concentrate (TC) is highly subjective and inconsistent, necessitating an objective, automated classification system. This study aims to develop a computationally efficient, low-cost model for TC quality classification using Histogram Features extracted from grayscale images combined with the K-Nearest Neighbor (KNN) algorithm. The methodology employed critical preprocessing steps, including StandardScaler for normalization and SMOTE for balancing the training data, followed by optimization across K=1 to K=30. The optimal model achieved a maximum accuracy of 69.23% at K=6, with an F1-Score of 71.43%, confirming robust performance on the imbalanced testing set. The results validate the effectiveness of the Histogram-KNN approach as a consistent and reliable decision support system for rapid TC quality screening in resource-limited settings.

Indonesian Sign Language (BISINDO) serves as a primary communication medium for the deaf community; however, limited public understanding often creates barriers during daily interactions. This study aims to develop a real-time BISINDO word-level translation system using hand landmark extraction and temporal modeling with Long Short-Term Memory (LSTM). The system employs MediaPipe Hands to detect 21 hand landmarks per frame, which are then processed as sequential motion patterns to classify five BISINDO words: saya, terima kasih, maaf, nama, and kamu. A total of 250 gesture samples were recorded under controlled lighting conditions as the primary dataset. The processed sequences were used to train the LSTM model, which was subsequently integrated with an ESP32 microcontroller and a DFPlayer Mini module to produce direct audio output. Experimental results show that the model achieved an average accuracy of 86%, with precision and recall values ranging from 0.81 to 0.94. The confusion matrix analysis indicates that most gestures were correctly classified, although some errors occurred in gestures with similar initial motion trajectories. Integration testing demonstrated an average system latency of 3.8 seconds and an audio output success rate of 85%. These findings indicate that the proposed system is capable of translating BISINDO word-level gestures accurately, responsively, and consistently in real-time conditions. This study provides a strong foundation for the broader development of sign language translation systems, with potential enhancements in vocabulary expansion, multi-user datasets, and hardware optimization for deployment in real-world environments.

Automatic Speech Recognition (ASR) for the Indonesian language faces significant challenges due to high Word Error Rate (WER), especially when using pre-trained models without fine-tuning. This study develops an optimized ASR system using a hybrid cloud architecture that integrates the Faster-Whisper large-v3 engine with advanced audio preprocessing techniques. The system adopts a distributed architecture, with Google Colab (Tesla T4, 15GB VRAM) as the GPU server and Ubuntu 22.04 LTS (8 core, 32GB RAM) as the client. Evaluation was conducted on five Indonesian audio samples covering formal news, informal conversations, and long-duration recordings. The system achieved an 80% success rate in processing, with WER ranging from 27.69% (formal news) to 645.16% (informal conversations). Resource utilization was also efficient, with 21.3% GPU usage and 35.4% RAM usage. Processing time remained stable for normal-sized files but experienced timeouts on large files (>50MB). The results indicate that hybrid cloud architecture is feasible for distributed ASR processing in Indonesian, with several areas still open for optimization toward production deployment.

Diabetes is a serious chronic disease that occurs because the pancreas does not produce enough insulin (a hormone that regulates blood sugar or glucose), or when the body cannot effectively use the insulin it produces. WHO data shows that the incidence of non-communicable diseases in 2004 reached 48 , 30% is slightly higher than the incidence rate of infectious diseases, namely 47.50% [1]. According to the Ministry of Health in 2012 diabetes caused 1.5 million deaths. Some Indonesian people, this disease is better known as diabetes or blood sugar. This research was developed through secondary data processing from the Pima Indians Diabetes Dataset health database which was taken from the Kaggle dataset and can be accessed through https://www.kaggle.com/uciml/pima-indians-diabetes-database. Where the data itself consists of 768 records with several medical predictor variables (Pregnancies, Glucose, Blood Pressure, Skin Thickness, Insulin, BMI, Diabetes Pedigree Function, Age and Outcome). Then the data will be processed using the Particle Swarm Optimization (PSO) feature selection to increase the accuracy value and the Naive Bayes algorithm to determine the accuracy results of the diagnosis of diabetes. From the results of research that has been done for the accuracy of the classification algorithm Naive Bayes is 74.61%, while the accuracy of the classification algorithm with Particle Swarm Optimization is 77.34% with an accuracy difference of 2.73%. So it can be concluded that the application of the Particle Swarm Optimization technique is able to select attributes in the Naive Bayes Algorithm, and can produce a better level of diabetes diagnosis accuracy than using only the individual method, namely the Naive Bayes algorithm. Keywords: Diabetes, Particle Swarm Optimization, Naive Bayes Algorithm