Claim Missing Document
Check
Articles

Found 3 Documents
Search
Journal : International Journal of Engineering, Science and Information Technology

Sentiment Analysis of Google Maps User Reviews on the Play Store Using Support Vector Machine and Latent Dirichlet Allocation Topic Modeling Zahrah, Violita Aditya; Nurdin, Nurdin; Risawandi, Risawandi
International Journal of Engineering, Science and Information Technology Vol 4, No 4 (2024)
Publisher : Department of Information Technology, Universitas Malikussaleh, Aceh Utara, Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52088/ijesty.v4i4.580

Abstract

These days, traveling is made easier by utilizing easily accessible online directions such as Google Maps. Google Maps provides real-time routes by displaying and presenting the closest routes that users can take. However, lately, the routes provided by Google Maps services often get users lost by presenting routes such as forests, narrow roads, and even dead ends. Therefore, this study aims to determine the level of user satisfaction and sentiment into two categories, namely positive and negative, based on reviews on the Google Play Store platform using the Support Vector Machine (SVM) algorithm and topic modeling using Latent Dirichlet Allocation (LDA) to find out the collection of topics that are the main topics of conversation by users regarding Google Maps services. The results of this study show that the SVM algorithm is feasible to use in sentiment analysis classification with an accuracy value of 86%, precision of 93%, recall of 53%, and f1-score of 52%. In addition, topic modeling is applied to generate coherence values for each topic, which shows that the higher the coherence value, the more specific the topic is. The highest coherence value generated in this study was two topic models with a coherence value of 35.15%, but this study took five with a coherence value of 33.39%. The five topic models to be applied in this study are selected because they have a good enough coherence value to identify the main topics and hidden topics in Google Maps user reviews with the Latent Dirichlet Allocation model. The topic model shows five aspects users often discuss: Google Maps route accuracy, system and service errors, navigation application directions, lost time history, and convoluted route provision.
Implementation of Organic and Inorganic Waste Selection System Based on Internet of Things Using MQTT Protocol at Abby Lhokseumawe Hospital Julita, Rina; Darnila, Eva; Risawandi, Risawandi
International Journal of Engineering, Science and Information Technology Vol 5, No 2 (2025)
Publisher : Malikussaleh University, Aceh, Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52088/ijesty.v5i2.826

Abstract

The waste sorting system designed for Abby Hospital in Lhokseumawe aims to improve the efficiency and effectiveness of waste management by automatically separating organic and inorganic materials. This system integrates Proximity sensors as the primary detectors, capable of detecting organic objects within a spatial range of 4 cm and inorganic objects within a range of 5 cm. The main feature of this system is its ability to automatically sort waste, which helps reduce the potential for human error in waste categorization and improve operational efficiency in the waste disposal process. During the testing phase, which focuses on assessing the trash bin's capacity when complete, the system uses ultrasonic sensors to measure and monitor the waste filling levels. The test results show an average data transmission delay of 445.33 ms, which is within the acceptable tolerance for this system. Additionally, the prototype is equipped with an operational status notification feature for users. This notification is delivered with an average delay of just 402.5 ms, ensuring that system status information is provided to users in real time. The combination of sensor detection precision and response speed in the waste sorting process highlights the system's effectiveness in improving waste management at the hospital. This system is expected to support the hospital's efforts in maintaining a clean environment and contribute to a more environmentally friendly and organized waste management program.
Data Mining Analysis for Clustering the Number of Tb Patients in North Aceh Health Centers Using the Spectral Method Clustering Khainesya, Khainesya; Darnila, Eva; Risawandi, Risawandi
International Journal of Engineering, Science and Information Technology Vol 5, No 2 (2025)
Publisher : Malikussaleh University, Aceh, Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52088/ijesty.v5i2.847

Abstract

Tuberculosis (TB) is one of the infectious diseases that is a significant concern in the world of health, especially in the North Aceh region. Grouping the number of TB patients based on severity and region is very important to support decision-making in further prevention and treatment efforts. This study applies the Spectral Clustering method to cluster the number of TB patients at Baktiya Health Center, Bayu Health Center, and Lhoksukon Health Center to identify patient distribution patterns based on severity categories. The system built is a web-based data mining analysis system using PHP and MySQL as a database. Clustering is done by dividing patients into three categories, low, medium, and high, based on five main criteria, namely age, gender, month of treatment, diagnosis results, and patient address. The results showed that Lhoksukon Health Center had the highest number of TB patients, with 136 patients (37.06%), an average age of 48.6 years, and the most cases occurred in December 2022. Bayu Health Center was at a moderate level with 130 patients (35.42%), most of whom were 45.5 years old, and most cases occurred in November 2023. Meanwhile, Baktiya Health Center had the lowest number of patients, 101 (27.52%), with the most cases occurring in November. From the clustering results, it can be concluded that the Spectral Clustering method can group TB patients well to help medical personnel and related parties develop more effective intervention strategies based on the region and severity of the patient.