International Journal of Advances in Data and Information System
Not a member yet
161 research outputs found
Sort by
Sentiment Analysis of the Sheikh Zayed Grand Mosque’s Visitor Reviews on Google Maps Using the VADER Method
Full text linkThe Sheikh Zayed Grand Mosque in Solo is a replica of the Zayed Grand Mosque in Abu Dhabi. Many people have provided reviews on Google Maps after visiting the mosque. This research aims to determine the sentiment results regarding visitors’ reviews by developing a sentiment analysis model using a combination of the Valance Aware Dictionary for Sentiment Reasoning (VADER) and Deep Translator methods. This research was conducted in two phases. The first phase proposed a sentiment analysis model using VADER and Deep-Translator with public datasets. Later, the resulting sentiment analysis model was applied in the second phase to analyze the dataset of mosque visitor reviews and determine public perceptions. This research compares two preprocessing models (PPTV1 and PPTV2) and continues with the translation and sentiment prediction processes. The evaluation results show the proposed model (PPTV2) achieved the best average accuracy values of 72%, precision of 83%, recall of 72%, and F1-Score of 75% for the three examined datasets. The results of visitor review sentiment obtained showed 83.3% positive, 9.5% neutral, and 7.2% negative. The analysis findings show that people are amazed by the beauty and majesty of the mosque. However, some people provide negative reviews of the mosque’s facilities
Machine Learning Algorithms for Prediction of Boiler Steam Production
Full text linkThe continuous increase in global electricity demand has resulted in boiler power plants becoming a significant energy source. The production of steam is a principal indicator of boiler efficiency, and the accurate prediction of steam production is paramount importance for the enhancement of boiler efficiency and the reduction of operational costs. In this study employs a boiler dataset with a steam production capacity of 420 tons per hour. A total of 25 independent variables were extracted from the original 39 variables through data processing and feature engineering for the purpose of prediction analysis. Subsequently, 8 machine learning models were used for modeling predictions. Grid search cross-validation was employed in order to optimise the performance of the model. The models were analysed and assessed using the Mean Squared Error (MSE) metrics. The results show that random forest achieves the highest accuracy among the 8 single models. Based on 8 models, New Bagging ensemble model is proposed, which combined predictions from 8 single models, demonstrated the optimal overall fit and the lowest MSE, achieved the purpose of the research. The present study demonstrates the ability to analyse and predict complex industrial systems with machine learning algorithms, and provides insights into the use of machine learning algorithms for industrial big data analytics and Industry 4.0. Further work could explore using larger datasets and deep learning to make predictions more accurate
Smart Expo UMKM Based on Extreme Programming Method: Evaluating on Black Box and UAT
Full text linkThe rapid development of information technology (IT) and the post-Covid-19 pandemic conditions provide joint learning to provide an online service that allows a business expo event to be carried out virtually and online based on the web. UMKM or Micro, small, and medium enterprises (MSMEs), as the spearhead of the people\u27s and regional economies, can utilize IT to improve product marketing and distribution through web-based virtual expos. For this reason, this study carried out the design and implementation of a web-based smart virtual expo using the Extreme Programming development method in a case study of the MSME business expo event in X Regency. The Extreme Programming development method was chosen to facilitate the adjustment of user needs to the virtual expo website developed through active communication with users during the development process and to make development possible in a relatively short time. This study uses a qualitative case study research method and the Black Box Testing method on the developer side and User Acceptance Testing (UAT) on the user side. The results of the study on Black Box Testing showed that all functionalities on the developed Smart Virtual Expo prototype can run well, and all users can accept the prototype well, with the largest age range at 41-50 years (56.8%) can use the prototype to access information and videos of products
Recommendation of Prospective Construction Service Providers in Government Procurement Using Decision Tree
Full text linkThe determination of prospective construction service providers using the direct procurement method is the authority of the Goods/ Services Procurement Officer. Administrative requirements are an important factor in selecting prospective construction service providers. The use of the decision tree method in this study is to find out, determine, and analyse the variables that influence the assessment of the feasibility of prospective construction service providers, and get an accuracy value in providing an assessment of the feasibility of prospective construction service providers. The data used in this study are 153 datasets consisting of 13 variables. The existing variables are divided into basic variables and additional variables. The basic variables consist of 5 variables, namely experts, work experience, quality of work, winning tenders and contract value. While the additional variables consist of 8 variables namely business entity status, business entity form, business entity NPWP, business entity domicile, business entity qualification, type of business licence, percentage of work and construction services business licence. By using the decision tree method, the accuracy on the basic variable is 84.84%. The addition of additional variables to the basic variables resulted in an accuracy of 90.91%. This shows that by adding additional variables the accuracy results are higher than using only the basic variables
Recommendation System for Selecting Web Programming Learning Materials for Vocational High School Students using Multi-criteria Recommendation Systems
Full text linkIn the independent curriculum, the learning that is carried out focuses on developing character, student competence and honing interests, talents. So the amount of learning material given to students does not have to be complete or less. Apart from that, the independent curriculum no longer burdens students with achieving a minimum score because assessments no longer use Minimum Completeness Criteria (KKM) scores. This makes it difficult for teachers to determine whether the material that has been explained can be understood because grades are not a benchmark for a student\u27s success. In fact, if the teacher does not know a student\u27s understanding, the teacher will have difficulty continuing to the next material. Implementation of the Multi-Criteria Recommender System (MCRS) can make it easier for teachers to predict whether students can progress to the next material and recommend which modules are suitable for these students. The recommendation system that will be built is in the form of web-based learning media so that students can be more interested and can help teachers improve learning outcomes. The method used is collaborative filtering by comparing adjusted cosine similarity, cosine based similarity and spearman rank order correlation. Based on the implementation of MCRS using the collaborative filtering method, it shows that the results of the recommendation system have a good impact on the teaching and learning process. Based on the 3 algorithms implemented, the best prediction result is cosine based similarity because the MAE value obtained is the lowest, namely 1.19 and the accuracy value is 76%
Exploring Sentiment Trends: Deep Learning Analysis of Social Media Reviews on Google Play Store by Netizens
Full text linkThis study explores sentiment analysis of Instagram app reviews using Long Short-Term Memory (LSTM) algorithms. The rise of app stores has transformed digital interactions, particularly for social media apps. Leveraging LSTM, we aim to understand user sentiments expressed in Instagram application reviews, offering insights to enhance user experience and address concerns. The methodology involves data crawling, preprocessing, LSTM model training, and evaluation metrics. Our findings reveal promising results in accurately identifying user sentiments, with an accuracy of 77.77%, precision of 0.45, recall of 0.089, and F1-score of 0.15. This study underscores the importance of sentiment analysis in understanding user feedback and its implications for app development and user engagement
Optimization of the Random Forest Method Using Principal Component Analysis to Predict House Prices: A Case Study of House Prices in Malang City
Full text linkInvestment is an interesting thing, especially property investment. The developer must also be careful in determining the price of the property. It should be noted that every year, both short-term and long-term, property prices increase and rarely go down. In determining the price, it is often also based on the features of the house such as the concept, location, bedrooms, etc. To predict house prices based on their features, the random forest has a good performance for predicting house prices. However, the random forest method has the disadvantage that if you use too many variables, the training process will take longer and feature selection tends to select features that are not informative. One way to reduce features without removing other features is to use Principal Component Analysis. In this research, the method used is Principal Component Analysis (PCA) and Random Forest. From the results of model training, it can be concluded that the use of model evaluation results using PCA has a smaller error rate and more consistent values, with an average of 0.018. While the results of the evaluation without PCA and using only Random Forest have a higher error value with an average of 0.03125. The training time using the PCA model has a faster time, with an average of 7918 milliseconds, while those using only random forest without PCA have an average time of 8975 milliseconds
Developing an Augmented Reality Application as Instruction Media to Help in Learning the Solar System
Full text linkThis study aimed to develop an application to help in learning the solar system using Augmented Reality (AR) technology. In an era where technology is very developed, learning about the solar system can be made more interactive by providing 3D illustrations to the students. One of the technologies that can be applied to support the development of educational applications to help in learning about the solar system is AR technology. It can create 3D illustrations. The study is the Research and Development (R&D). The research produced an AR-based solar system introduction application. This application can be used as a learning media for students. The developed AR application was tested using alpha and beta testing. The alpha testing was the marker accuracy testing and black-box testing, while the beta testing was done by distributing questionnaires to 30 respondents and then doing validity and reliability test. This study produced an AR application to help in learning the solar system. The black-box testing showed that the AR application generally was functioning well. The marker accuracy testing showed that the AR camera succeeded in scanning markers up to 25% of the marker area. The data obtained from distributing questionnaires were processed to know the validity in terms of attractiveness and effectiveness, and the results showed the data was valid. Moreover, the reliability testing was carried out with Cronbach\u27s alpha, and the result was 0.771 for the attractiveness aspect and 0.742 for the effectiveness aspect. These values mean that most beta testers agree that the AR application was attractive and effective
Implementation of Random Search Algorithm with FSSRS (Fixed Step Size Random Search) for Applicating the Patrol System Based on Mobile Computing
Full text linkEnvironmental security is very influential for the sustainability of human life. In order for environmental security to remain in a safe condition, a system is needed that can control the environment, such as patrolling at every point to ensure that the environmental conditions are safe. However, it is felt that this is not enough if the patrol system is not assisted by tools or systems that are digitalized and integrated with community service officers, such as firefighters, ambulances, and police, and are easy for officers to use when conducting patrols. So, it is necessary to schedule patrols to several points with different routes for each activity so that it is not easily read by unwanted parties in terms of crime. In order for the system to obtain patrol scheduling in a timely and efficient manner, an appropriate and efficient algorithm is needed, the algorithm is random search with FSSRS (Fixed Step Size Random Search) which can suggest random and precise patrol scheduling. From the results of training using four iterations, namely 50, 100, 150, and 200, the best value was produced in the 200th iteration. Data was taken from the results of a case study survey with eight patrol points using coordinates at each point. So, it can be concluded that the FSSRS algorithm is effectively used to randomize patrol points and can be implemented in the application patrol system
The Construction of Counseling Information System with Object Oriented Technology Approach
Full text linkThe development of information systems in currently preserve to adjustment and it was customized according to the problem domain. In this paper, we will describe the modeling language that is applied when developing an information system. The development of information systems has become a trend for the last decade. Various methods have been proposed with different characteristics. The counseling information system in this case was developed with an object-oriented approach. The strong motivation in this research is to answer the challenges and demands of system development. Modeling in system development explains various modeling languages starting with identifying problems with use case modeling languages (Use case Diagrams), simplifying problems with structured modeling (Class Diagrams), to building communication patterns or interactions between objects with one another (Sequence Diagram). Thus, the system built is able to describe the complexity of information systems, especially counseling information