Face mask detection is a technological application that employs computer vision methodologies to ascertain the presence or absence of a face mask on an individual depicted in an image or video. This technology gained significant attention and adoption during the COVID-19 pandemic, as wearing face masks became an important measure to prevent the spread of the virus. Face mask detection helps to enforce mask-wearing guidelines, which can significantly reduce the spread of respiratory illnesses, including COVID-19. Wearing masks in densely populated areas provides individuals with protection and hinders the spread of airborne particles that transmit viruses. The application of deep learning models in object recognition has shown significant progress, leading to promising outcomes in the identification and localization of objects within images. The primary aim of this study is to annotate and classify face mask entities depicted in authentic images. To mitigate the spread of COVID-19 within public settings, individuals can employ the use of face masks created from materials specifically designed for medical purposes. This study utilizes YOLOv8, a state-of-the-art object detection algorithm, to accurately detect and identify face masks. To analyze this study, we conducted an experiment in which we combined the Face Mask Dataset (FMD) and the Medical Mask Dataset (MMD) into a single dataset. The detection performance of an earlier research study using the FMD and MMD was improved by the suggested model to a “Good” level of 99.1%, up from 98.6%. Our study demonstrates that the model scheme we have provided is a reliable method for detecting faces that are obscured by medical masks. Additionally, after the completion of the study, a comparative analysis was conducted to examine the findings in conjunction with those of related research. The proposed detector demonstrated superior performance compared to previous research in terms of both accuracy and precision.

Pengunjung merupakan hal yang penting bagi Perpustakaan. Untuk mendapatkan pengunjung yang banyak diperlukan layanan yang tersedia bagi pengunjung. Perpustakaan dan Arsip Daerah Salatiga (Persipda) adalah salah satu perpustakaan di Salatiga yang memiliki banyak pengunjung dan menyediakan berbagai layanan pengunjung. Data pelayanan pengunjung merupakan hal yang penting untuk mengetahui tingkat kualitas pelayanan. Kualitas pelayanan dapat diketahui dengan terjadinya peningkatan atau penurunan jumlah pengguna suatu layanan. Oleh karena itu dibutuhkan sistem informasi data pelayanan pengunjung untuk mengelola data di pelayanan serta mengetahui terjadinya peningkatan atau penurunan jumlah pelayanan pengunjung. Sistem informasi ini dilengkapi fitur highchart untuk menampilkan data pengunjung dalam bentuk chart yang ada di database sistem.

This research focuses on the development of the Plant Encyclopedia REST API, a website application designed using Java, HTML, Spring Boot, MyBatis Generator, and Swagger technologies, with PostgreSQL as the database used to store plant data from the API. This study highlights the declining abilities of the younger generation, especially in mathematics, literacy, and science. Statistical data from PISA test scores from 2003 to 2022 on 15-year-olds worldwide show a peak decline in 2022. The applied research methods include needs analysis, problem identification, system design, application design, implementation, and system testing. This research aims to overcome the lack of knowledge about plants that can have a negative impact on the ecosystem and provide various other benefits. The result of this research is a plant encyclopedia website application that can be accessed by various users to find information related to plants and a swagger website for administrators to manage plant data.

ANALISIS CENTRALITY DAN SENTIMENT PERCAKAPAN TWITTER(X) TERKAIT GIBRAN MENGGUNAKAN SNA DAN VADER

The Salatiga Regional General Hospital (RSUD) plays an important role in providing healthcare services. This research analyzes public sentiment towards RSUD Salatiga using the SVM method with a linear kernel for sentiment analysis and TF-IDF for feature extraction. The dataset consists of 414 processed reviews, including case folding, data cleaning, tokenization, normalization, stopword removal, and stemming. Evaluation shows that the model achieved an accuracy of 84.00%; precision of 84.00%; recall of 83.25%; and an F1-score of 83.53%. A total of 55.8% of reviews indicated positive sentiment and 44.2% negative sentiment, highlighting the need for improvements in the queue system, waiting times, and parking facilities. The SVM and TF-IDF methods were chosen for their ability to handle large text data with high accuracy. This research provides practical contributions in the form of recommendations such as the implementation of a technology-based queue system. Limitations include the limited amount of data and platform bias, so exploring other algorithms, such as Naive Bayes and Random Forest, is recommended.