Search

Background: Hepatitis is a contagious inflammatory disease of the liver and is a public health problem because it is easily transmitted. The main factors causing hepatitis are viral infections, disease complications, alcohol, autoimmune diseases, and drug effects. Some hepatitis variants such as B, C, and D can also cause liver cancer if left untreated. Objective: This research aims to determine the effect of Backward Elimination feature selection on the performance of hepatitis disease identification compared to cases where Backward Elimination is not applied. Methods: XGBoost classification, capable of handling machine learning problems, was utilized. Additionally, Backward Elimination was used as a featured selection to increase accuracy by reducing the number of less important features in the data classification process. Results: The results for training XGBoost model with Backward Elimination, and applying Random Search for hyperparameter optimization, achieved an accuracy of 98.958% at 0.64 seconds. This performance was better than using Bayesian search, which produced the same accuracy of 98.958% but required a longer training time of 0.70 seconds. Conclusion: The use of features obtained from Backward Elimination process as well as the use of feature average values for missing value treatment, produced an accuracy of 98.958%.the precision in training XGBoost model with hyperparameter Bayesian search achieved accuracy, recall, and F1 score of 98.934%, 98.934%, and 98.934%, respectively. Consequently, the use of Backward Elimination in XGBoost model led to faster training, improved accuracy, and decreased overfitting. Keywords: Hepatitis, Backward Elimination, XGBoost, Bayesian Search, Random Search

Background: Stemming is significantly essential in natural language processing (NLP) due to the ability to minimize word variations to fundamental forms. This procedure facilitates the analysis of textual data and enhances the precision of classification and information retrieval. Objective: Previous related systematic literature review has not been conducted on stemming and lemmatization in regional languages in Indonesia. Therefore, this study aims to conduct a systematic literature review to capture the latest developments in stemming and lemmatization in regional languages in Indonesia. Methods: This study was carried out using Kitchenham method, analyzing 35 studies extracted from 740, which were obtained from Scopus, IEEE Xplore, and Google Scholar, and published between 2014 and 2023. Results: The results showed that study trends in stemming possessed the potential to continue developing every year. Additionally, the main element in stemming and lemmatization studies was found to be the availability of digital dictionaries in regional languages. This was because greater number of basic vocabularies contributed more positively to stemming or lemmatization. The availability of word morphology information in regional languages would be constructive for making rule-based stemmers. Meanwhile, corpus-based stemming and lemmatization studies could only be conducted for languages with a large corpus to ensure there were various affixed words to process. Conclusion: Based on SLR study, stemming and lemmatization in regional languages in Indonesia developed significantly from 2014 to 2023. The two main strategies applied included using available digital dictionaries and language morphology information. However, the main challenges encountered were the limited number of vocabulary words in the dictionaries and testing various rule-based methods. Keywords: Lemmatization, Morphology, Rule-based, Stemming, Systematic Literature Review.

Background: The most commonly used mathematical model for analyzing disease spread is the Susceptible-Exposed-Infected-Recovered (SEIR) model. Moreover, the dynamics of the SEIR model depend on several factors, such as the parameter values. Objective: This study aimed to compare two optimization methods, namely genetic algorithm (GA) and particle swarm optimization (PSO), in estimating the SEIR model parameter values, such as the infection, transition, recovery, and death rates. Methods: GA and PSO algorithms were compared to estimate parameter values of the SEIR model. The fitness value was calculated from the error between the actual data of cumulative positive COVID-19 cases and the numerical data of cases from the solution of the SEIR COVID-19 model. Furthermore, the numerical solution of the COVID-19 model was calculated using the fourth-order Runge-Kutta algorithm (RK-4), while the actual data were obtained from the cumulative dataset of positive COVID-19 cases in the province of Jakarta, Indonesia. Two datasets were then used to compare the success of each algorithm, namely, Dataset 1, representing the initial interval for the spread of COVID-19, and Dataset 2, representing an interval where there was a high increase in COVID-19 cases. Results: Four parameters were estimated, namely the infection rate, transition rate, recovery rate, and death rate, due to disease. In Dataset 1, the smallest error of GA method, namely 8.9%, occurred when the value of , while the numerical error of PSO was 7.5%. In Dataset 2, the smallest error of GA method, namely 31.21%, occurred when , while the numerical error of PSO was 3.46%. Conclusion: Based on the parameter estimation results for Datasets 1 and 2, PSO had better fitting results than GA. This showed PSO was more robust to the provided datasets and could better adapt to the trends of the COVID-19 epidemic. Keywords: Genetic algorithm, Particle swarm optimization, SEIR model, COVID-19, Parameter estimation.

Background: Internet topology is studied more by researchers on global internet coverage compared to limited regional coverage. However, some studies also see the importance of studying internet topology in certain countries or regions. The internet performance of a country or region can be influenced by the structure of its internet topology, and research on internet topology can contribute to improving internet topology in that region. Objective: This research initially carried out internet topology mapping in a limited region, then used experience from each step in conducting internet topology mapping to create a comprehensive guide on procedures for internet topology mapping at the autonomous system (AS) level in certain regional areas. Methods: Internet topology mapping is carried out by inferring relationships between ASes through an inference process against border gateway protocol (BGP) table dumps, while the internet topology mapping method chosen is passive mapping. Results: The entire series of steps involved in the regional internet topology mapping process have been successfully outlined in a detailed guide as a result of this research. Evaluation of the research results was carried out by implementing the application of this comprehensive guide and also through assessments from experts in related fields regarding the results of this research. The results of both evaluations showed that the research results were appropriate. Conclusion: This research provides a comprehensive guide for mapping internet topology in specific regional areas, consisting of nine sequential steps grouped into four major steps. This guide can be used to assist similar research efforts in other regional areas as well as provide further knowledge regarding studies in this field. This research is different from previous studies, because it provides a comprehensive guide to the internet topology mapping process, which has not been available in previous studies. Keywords: Internet Topology Mapping, Regional Internet Topology, Autonomous System, Border Gateway Protocol

Background: To understand land transformation at the local level, there is a need to develop new strategies appropriate for land management policies and practices. In various geographical research, ground coverage plays an important role particularly in planning, physical geography explorations, environmental analysis, and sustainable planning. Objective: The research aimed to analyze land cover using vegetation density data collected through remote sensing. Specifically, the data assisted in land processing and land cover classification based on vegetation density. Methods: Before classification, image was preprocessed using Convolutional Neural Network (CNN) architecture's ResNet 50 and DenseNet 121 feature extraction methods. Furthermore, several algorithm were used, namely Decision Tree, Naí¯ve Bayes, K-Nearest Neighbor, Random Forest, Support Vector Machine (SVM), and eXtreme Gradient Boosting (XGBoost). Results: Classification comparison between methods showed that using CNN method obtained better results than machine learning. By using CNN architecture for feature extraction, SVM method, which adopted ResNet-50 for feature extraction, achieved an impressive accuracy of 85%. Similarly using SVM method with DenseNet121 feature extraction led to a performance of 81%. Conclusion: Based on results comparing CNN and machine learning, ResNet 50 architecture performed the best, achieving a result of 92%. Meanwhile, SVM performed better than other machine learning method, achieving an 84% accuracy rate with ResNet-50 feature extraction. XGBoost came next, with an 82% accuracy rate using the same ResNet-50 feature extraction. Finally, SVM and XGBoost produced the best results for feature extraction using DenseNet-121, with an accuracy rate of 81%. Keywords: Classification, CNN Architecture, Feature Extraction, Ground Coverage, Vegetation Density.

Background: Motorcycle taxi (MCT) is a form of widely used informal transportation. Over the last few years, MCT has developed from conventional public transportation into a modern business using digital technology. In these services, digitalization has received a positive response from the public and the increasing number creates many challenges. However, there is a gap in the discussion of new service models for shared mobility transportation. Objective: This research aimed to analyze MCT, providing an overview of the services as informal transportation. Methods: Bibliometric analysis was used to evaluate 366 articles published in the Scopus database between 2011-2023. In addition, annual publications and citation topics, most productive sources and influential articles, relevant affiliations, productive countries, location research, main topics, and future research options were identified. Results: Frequently occurring topics were reported with past and present academic developments related to MCT services. Based on the publication themes, the main topics were arranged into five clusters, namely (a) the development of more sustainable transportation services, (b) environmental and health impact, (c) road safety, (d) risky behaviors or risk factors as MCT drivers, and (e) utilization of MCT for medical or health services. Meanwhile, the research topics comprised themes about travel behavior, health, safety-security, customer satisfaction, and advanced mobility topics. Conclusion: This research increased knowledge about main topics, trends and future analysis options in MCT. The academic developments served as a guide to future topics. Keywords: Motorcycle taxi, public transportation, informal transportation, shared mobility, bibliometric analysis

Background: Plant diseases affect agricultural output, quality and profitability, making them serious obstacles for agriculture. It is essential to detect diseases early in order to reduce losses while retaining sustainable practices. Plant disease detection has benefited greatly from the use of computer vision and deep learning in recent years because of their outstanding precision and computing capability. Objective: In this paper, we intend to investigate the role of deep learning in computer vision for plant disease detection while looking into how these techniques address complex disease identification problems. A variety of deep learning architectures were reviewed, and the contribution of frameworks such as Tensorflow, Keras, Caffe and PyTorch to the researchers' model construction was studied as well. Additionally, the usage of open repositories such as PlantVillage and Kaggle along with the customized datasets were discussed. Methods: We gathered the most recent developments in deep learning techniques for leaf disease detection through a systematic literature review of research papers published over the past decade, using reputable academic databases like Scopus and Web of Science, following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) method for guidance. Results: This study finds that researchers consistently enhance existing deep learning architectures to improve prediction accuracy in plant disease detection, often by introducing novel architectures and employing transfer learning methods. Frameworks like TensorFlow, Keras, Caffe, and PyTorch are widely favored for their efficiency in development. Additionally, most studies opt for public datasets such as PlantVillage, Kaggle, and ImageNet, which offer an abundance of labelled data for training and testing deep learning models. Conclusion: While no singular ‘best' model emerges, the adaptability of deep learning and computer vision demonstrates the dynamic nature of plant disease recognition area, and this paper provides a comprehensive overview of deep learning's transformative impact on plant disease recognition by bringing together information from different studies. Keywords: Deep learning, Computer vision, Plant disease, Systematic literature review

Background: Technology Acceptance Model (TAM) and Expectation-Confirmation Model (ECM) integration model are commonly used to analyze the intention to use technology in education. Moreover, the ease of implementation causes various external factors influencing technology acceptance to continue growing. However, limited research focuses on the use of TAM and ECM in the acceptance of cloud-based academic system. Objective: This research aims to identify factors influencing user perceptions of cloud-based academic information system and the relationships among different factors. Methods: The research integrated Extended TAM and ECM, subsequently processing data obtained from 261 respondents using Structural Equation Modeling-Partial Least Squares (SEM-PLS). The perceptions proposed included Facilitating Condition (FC), Perceived Usefulness (PU), Perceived Ease of Use (PEOU), Confirmation (CM), Satisfaction (SF), and Behavioral Intention to Use (BIU). Results: Based on the data processing carried out, the results were PEOU against BIU (H1, êžµ=0.256, p=0.001), PU against BIU (H2, êžµ=0.200, p=0.007), and SF against BIU (H3, êžµ=0.499, p= 0.000). Furthermore, it also comprised FC against PEOU (H4, êžµ=0.839, p=0.000), PU (H5, êžµ=0.849, p=0.000) and SF (H6, êžµ=0.294, p=0.000), as well as CM against SF (H7, êžµ=0.358, p=0.000) and PU against SF (H8, êžµ=0.325, p=0.000). These results showed that each proposed construct significantly influenced behavioral intentions to use cloud-based academic information system. Conclusion: The results showed that each factor proposed in the construct significantly influenced user intentions to use cloud-based academic system. Consequently, the most influential drivers in using cloud-based academic system were SF, PU, PEOU, and FC. Keywords: Acceptance, Behavioral Intention, Cloud-Based Academic System, Expectation

Background: Amidst a rapidly evolving digital landscape that accelerates the flow of information, higher education institutions face the unique challenge of managing vast and dynamic knowledge resources. This research delves into the motivations and innovative solutions for developing Knowledge Management Systems (KMS), which is key to optimizing knowledge resource utilization and enhancing academic collaboration. Objective: This research provides a comprehensive mapping of problems and solutions for developing university knowledge management systems based on previous research. Not only that, but the results of this study also suggest three future research studies that can be adopted. Methods: This study used the Kitchenham systematic literature review method. The author uses literature in the form of journals and conference proceedings published from 2019 to 2023. Twenty-three articles were used for this study from 5 databases, such as ACM, ProQuest, Scopus, Taylor & Francis, and IEEE Xplore. Results: This study reveals research trends in knowledge management systems within higher education, examining aspects such as country, data collection methods, research methodologies, and theoretical frameworks. The main problems motivating the development of KMS are identified and categorized based on the people, process, and technology framework. In overcoming these problems in the university business process, there are several alternative solutions, both in the form of requirements and systems. Thus, the results of this study seek to provide guidelines for future research to adopt alternative solutions from this research and develop KMS to provide new solutions. Conclusion: This study advances knowledge about various trends, motivations, requirements, and system solutions to address KMS problems in higher education. The authors' research results can add valuable insights to improve our understanding of the development of KMS in universities in various countries. Future research can identify new potential in KMS in business processes currently running in a university with appropriate methodologies. Keywords: Knowledge management system, higher education, systematic literature review, problem, solution

Background: In Agile software development, an essential initial stage is eliciting software requirements. This process engages stakeholders to achieve comprehensive results. However, a common issue is the variance in domain and technical knowledge among stakeholders, potentially impacting the quality of software requirements elicitation. Objective: Understanding the characteristics of user stories produced by stakeholders becomes crucial, particularly considering the differences in domain and technical knowledge. This study aims to compare the characteristics of user stories generated by stakeholders with varying backgrounds in domain and technical expertise. Methods: The initial step involves categorizing respondents into distinct stakeholder groups. Three stakeholders are involved in this study, constituting a combination of those with high and low technical and domain knowledge. Subsequently, data collection of user stories is conducted across various case studies. Finally, the acquired user stories are analyzed for further insights. Results: The analysis reveals variations in user stories generated by the three stakeholder categories across the three case studies. Stakeholders with domain knowledge tend to focus on 'what' aspects with task elements and 'why' aspects with hard-goal elements. Meanwhile, technical knowledge crafts user stories with capability elements in the 'what' aspect. Utilizing the QUS framework, it is evident that technical knowledge consistently produces a higher number of high-quality user stories across all quality categories, Conclusion: The contribution offered by this study lies in determining the distinct characteristics of user stories produced by different types of stakeholders, focusing on disparities in domain and technical knowledge. The study highlights the comparison of various characteristics of user story elements, such as hard-goals, soft-goals, tasks, or capabilities, and assesses the quality of user stories based on the user story framework. Additionally, it endorse the importance of process innovation in shaping the requirements gathering process and subsequently influencing the quality of user stories. Keywords: User story, Agile Software Development, Requirements Elicitation, Stakeholder, Domain Knowledge, Process Innovation

Background: The widespread use of financial technology (FinTech) is a popular aspect across various fields, particularly in online transportation. However, the usage has led to an increase in illegal FinTech, causing significant problems for public. Issues related to account security, such as hacks leading to the loss of user balances and misuse of data, contribute to the erosion of brand image and public trust. Despite the growing prominence of FinTech, explorations on the application in the context of online transportation remain limited. Previous studies have not discussed the impact of brand image on perceived usefulness and ease of use. Therefore, this current study explores the importance of combining brand image and trust factors to increase user intention. This process is achieved by investigating brand image and trust as crucial factors influencing increased perceived ease and benefits during the integration of FinTech in online transportation services. Objective: This study aimed to measure the impact of brand image and trust factors on the adoption of FinTech in online transportation. Methods: The investigation was carried out with a quantitative analysis approach using Partial Least Squares–Structural Equation Modeling (PLS–SEM). Furthermore, it focused on understanding FinTech services in online transportation, incorporating factors such as trust, brand image, perceived ease of use, perceived usefulness, and user intention. Data were collected by using a purposive sampling method through online questionnaire distribution. PLS-SEM was adapted for analyzing variable relationships, hypotheses, and models. Results: The results showed that factors including trust, perceived ease of use, and perceived usefulness significantly influenced the willingness to use FinTech in online transportation services. However, it was observed that brand image factors did not impact user intentions. Conclusion: This study showed a critical aspect in understanding the value of FinTech services by explaining the importance of establishing trust and building a good brand image as precursors. These factors indirectly contributed to increased perceived benefits and ease of use. Therefore, the insights offered valuable input for companies aiming to develop trusted FinTech platforms with a positive product image. Keywords: Brand Image, Trust, FinTech, Online Transportation

Background: This study focuses on the latest knowledge regarding Micro, Small and Medium Enterprises (MSMEs) as a current central issue. These enterprises have shown their significance in providing employment opportunities and contributing to the country's economy. However, MSMEs face various challenges that must be addressed to optimize their outcomes. Understanding the characteristics of this group was crucial in formulating effective strategies. Objective: This study proposed to cluster or combine micro, small, and medium enterprises (MSMEs) data in a particular area based on asset value and turnover. As a result, this study aimed to gain insights into the MSME landscape in the area and provided valuable information for decision-makers and stakeholders. Methods: This study utilized two methods, namely the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) method and the K-Means method. These methods were chosen for their distinct capabilities. DBSCAN was selected for its ability to handle noisy data and identify clusters with diverse forms, while K-Means was chosen for its popularity and ability to group data based on proximity. The study used a dataset containing MSME information, including asset values and turnover, collected from various sources. Results: The outcomes encompassed identifying clusters of MSMEs based on their closeness in the feature space within a specific region. Optimizing the clustering outcomes involved modifying algorithm parameters like epsilon and minimum points for DBSCAN and the number of clusters for K-Means. Furthermore, this study attained a deeper understanding of the arrangement and characteristics of MSME clusters in the region through a comparative analysis of the two methodologies. Conclusion: This study offered perspectives on clustering MSMEs based on asset value and turnover in a specific region. Employing DBSCAN and K-Means methodologies allowed researchers to depict the MSME landscape and grasp the business attributes of these enterprises. These results could aid in decision-making and strategic planning concerning the advancement of the MSME sector in the mentioned area. Future study may investigate supplementary factors and variables to deepen comprehension of MSME clusters and promote regional growth and sustainability. Keywords: Asset Value, Clustering, DBSCAN, K-Means, Turnover

Background: Insect vectors spread 80% of plant viruses, causing major agricultural production losses. Direct insect vector identification is difficult due to a wide range of hosts, limited detection methods, and high PCR costs and expertise. Currently, a biodiversity database named Global Biotic Interaction (GloBI) provides an opportunity to identify virus vectors using its data. Objective: This study aims to build an insect vector search engine that can construct an virus-insect-plant interaction knowledge graph, identify insect vectors using network analysis, and extend knowledge about identified insect vectors. Methods: We leverage GloBI data to construct a graph that shows the complex relationships between insects, viruses, and plants. We identify insect vectors using interaction analysis and taxonomy analysis, then combine them into a final score. In interaction analysis, we propose Targeted Node Centric-Degree Centrality (TNC-DC) which finds insects with many directly and indirectly connections to the virus. Finally, we integrate Wikidata, DBPedia, and NCBIOntology to provide comprehensive information about insect vectors in the knowledge extension stage. Results: The interaction graph for each test virus was created. At the test stage, interaction and taxonomic analysis achieved 0.80 precision. TNC-DC succeeded in overcoming the failure of the original degree centrality which always got bees in the prediction results. During knowledge extension stage, we succeeded in finding the natural enemy of the Bemisia Tabaci (an insect vector of Pepper Yellow Leaf Curl Virus). Furthermore, an insect vector search engine is developed. The search engine provides network analysis insights, insect vector common names, photos, descriptions, natural enemies, other species, and relevant publications about the predicted insect vector. Conclusion: An insect vector search engine correctly identified virus vectors using GloBI data, TNC-DC, and entity embedding. Average precision was 0.80 in precision tests. There is a note that some insects are best in the first-to-five order. Keywords: Knowledge Graph, Network Analysis, Degree Centrality, Entity Embedding, Insect Vector

Background: The publicly available large dataset plays an important role in the development of the natural language processing/computational linguistic research field. However, up to now, there are only a few large Indonesian language datasets accessible for research purposes, including sentiment analysis datasets, where sentiment analysis is considered the most popular task. Objective: The objective of this work is to present sentiment analysis on a large Indonesian product review dataset, employing various features and methods. Two tasks have been implemented: classifying reviews into three classes (positive, negative, neutral), and predicting ratings. Methods: Sentiment analysis was conducted on the FDReview dataset, comprising over 700,000 reviews. The analysis treated sentiment as a classification problem, employing the following methods: Multinomial Naí¯ve Bayes (MNB), Support Vector Machine (SVM), LSTM, and BiLSTM. Result: The experimental results indicate that in the comparison of performance using conventional methods, MNB outperformed SVM in rating prediction, whereas SVM exhibited better performance in the review classification task. Additionally, the results demonstrate that the BiLSTM method outperformed all other methods in both tasks. Furthermore, this study includes experiments conducted on balanced and unbalanced small-sized sample datasets. Conclusion: Analysis of the experimental results revealed that the deep learning-based method performed better only in the large dataset setting. Results from the small balanced dataset indicate that conventional machine learning methods exhibit competitive performance compared to deep learning approaches. Keywords: Indonesian review dataset, Large dataset, Rating prediction, Sentiment analysis

Background: Over the years, the role of startups has experienced a significant increase in healthcare delivery, particularly in telemedicine. However, there are still some inherent challenges, including cultural factors, lack of digital literacy, and uneven internet network infrastructure that must be considered during implementation. Previous reports also showed that there was a knowledge gap regarding the factors influencing acceptance of telemedicine. Objective: This study aimed to introduce and investigate an adjusted model based on Technology Acceptance Model (TAM) to assess the influence of user dimensions, technological aspects, and socio-cultural elements on the intention to adopt telemedicine services. Methods: The hypothesized relationships between latent variables were examined through Structural Equation Modeling (SEM). In addition, data analysis was carried out using Partial Least Squares-Structural Equation Modeling (PLS-SEM). Results: Self-efficacy (β=-0.272, P=0.013), perceived usefulness (β=0.355, P=0.000), facilitating conditions (β=0.425, P=0.000), and cultural factors (β=0.421, P=0.001) were found to exert a significant influence on the intention to adopt telemedicine services. Meanwhile, trust, the variables of perceived ease of use, and social influence had no significant influences. Conclusion: This study emphasized the significance of comprehending the factors influencing the adoption of telemedicine services. In addition, the results showed that the extended TAM was applicable in assessing acceptance of telemedicine services. Keywords: acceptance, telemedicine, TAM, SEM, intention to use

Background: Mobile social commerce is a collection of e-commerce activities accessed via mobile devices and supported by users actively engaging in commercial activities on social media. As a country with a substantial number of social media users, Indonesia has sufficient opportunities to implement mobile social commerce as application for online shopping. Objective: This study aimed to identify the factors influencing the use of mobile social commerce for online shopping, using Unified Theory of Acceptance and Use of Technology 2 (UTAUT2). In this context, some variables were excluded, namely user behavior, price value, and moderating variables (age, gender, and experience). Additional variables considered included price saving orientation (PSO), privacy concerns (PC), social commerce construct (SCC), social support (SS), and trust (TR). Methods: Data were collected by distributing questionnaires to respondents who had used mobile social commerce for shopping, resulting in 320 collected responses. Furthermore, the collected data were analyzed using Partial Least Square-Structural Equation Modeling (PLS-SEM) method through SmartPLS 3.3.3 application. Results: The results showed that among the 17 proposed hypotheses, 6 were rejected, while 11 were accepted. Conclusion: In conclusion, the factors influencing the use of mobile social commerce consisted of effort expectancy, habit, hedonic motivation, SCC, SS, and PC. Therefore, future studies should concentrate on exploring the continued intention of users towards mobile social commerce application. Keywords: Mobile Social Commerce, Privacy Concern, Social Construct, Social Support, UTAUT

Background: Parkinson's disease (PD) is a critical neurodegenerative disorder affecting the central nervous system and often causing impaired movement and cognitive function in patients. In addition, its diagnosis in the early stages requires a complex and time-consuming process because all existing tests such as electroencephalography or blood examinations lack effectiveness and accuracy. Several studies explored PD prediction using sound, with a specific focus on the development of classification models to enhance accuracy. The majority of these neglected crucial aspects including feature extraction and proper parameter tuning, leading to low accuracy. Objective: This study aims to optimize performance of voice-based PD prediction through feature extraction, with the goal of reducing data dimensions and improving model computational efficiency. Additionally, appropriate parameters will be selected for enhancement of the ability of the model to identify both PD cases and healthy individuals. Methods: The proposed new model applied an OpenML dataset comprising voice recordings from 31 individuals, namely 23 PD patients and 8 healthy participants. The experimental process included the initial use of the SVM algorithm, followed by implementing PCA for feature extraction to enhance machine learning accuracy. Subsequently, data balancing with SMOTE was conducted, and GridSearchCV was used to identify the best parameter combination based on the predicted model characteristics. Result: Evaluation of the proposed model showed an impressive accuracy of 97.44%, sensitivity of 100%, and specificity of 85.71%. This excellent result was achieved with a limited dataset and a 10-fold cross-validation tuning, rendering the model sensitive to the training data. Conclusion: This study successfully enhanced the prediction model accuracy through the SVM+PCA+GridSearchCV+CV method. However, future investigations should consider an appropriate number of folds for a small dataset, explore alternative cross-validation methods, and expand the dataset to enhance model generalizability. Keywords: GridSearchCV, Parkinson Disaese, SVM, PCA, SMOTE, Voice/Speech

Background: Maternal and child health in Integrated Healthcare Center Information Systems (IHCIS) has been implemented in several community health centers. Some have been implemented successfully, but others have failed. Many factors influence the success and failure of IHCIS implementation. Thus, knowing these factors can be used to assist the decision-making process in implementing IHCIS. Objective: This research aims to determine the factors affecting the success and failure of IHCIS for maternal and child health using the HOT-Fit (Human, Organization, Technology, and Fit) model. Methods: This research begins with preliminary research to identify the problem, determine research variables, and collect data. It uses quantitative and qualitative approaches. A quantitative approach is conducted at locations that have successfully implemented IHCIS. The data collection instrument uses a questionnaire. A qualitative approach was conducted in locations that were still experiencing failure in implementing IHCIS. Data collection techniques through direct interviews. Results: Organizational factors do not fully determine the success or failure of the information system. Factors supporting the success of IHCIS are human (user satisfaction and system use) and technological factors (quality of information and the quality of service). However, technology (system quality and information quality) is the main factor in the failure of IHCIS implementation. Problems with system quality include the system login, limited access to the internet, and an information system that is not in accordance with requirements. The perceived information obstacle is that the system is not yet integrated, and the information produced is incomplete. Conclusion: To satisfy requirements, the information and system qualities must be enhanced. Implementing IHCIS requires an appropriate strategy according to local circumstances and conditions. This approach involves human, organizational, and technological factors. Keywords: Community Health Workers, Evaluation, HOT-Fit, Integrated Healthcare Center Information Systems, Success Factors

Background: The exponential data growth emphasises the importance of efficient information flow in organisations, especially in the financial sector. Data quality significantly influences decision-making, necessitating reliable Data Quality Management (DQM) frameworks. Previous studies propose DQM to maintain data quality through regulation, technology, measurement, evaluation, and improvement. Researchers highlight high-quality data benefits in private organisations but note the lack of improvement in data utilisation in public organisations. In Indonesia, data accuracy and quality are crucial for financial policies, with frequent reports of data inaccuracies in the Directorate General of Customs and Excise (DJBC), demanding standardised DQM practices. However, However, prior studies have yet to provide comprehensive and practical solutions to improve DQM practices. This study therefore aims to measure the DQM maturity, provide recommendations based on best practices, and formulate a practical strategy for improvements along with indicators tailored to the organisation, a topic that previous research has not explored. Methods: This study falls under a mixed method approach (a quantitative study followed by a qualitative study) and employs a three-stage methodology. The authors conduct maturity assessment using Loshin model through an assisted enumeration from 5 key stakeholders followed by recommendations based on the Data Management Body of Knowledge (DMBOK) and strategy formulation from internal documents and interview. Results: The data analysis yielded a DQM maturity score of 3.10, indicating a "defined to managed" level of maturity. Among eight components, only one receives a Managed level, two components are in the Defined level and the rest belongs to a Repeatable level. This study also proposes three strategies to bolster DQM by targeting 49 weak points, which will be progressively and sequentially implemented over a three-year period, using twelve possible solutions. Conclusion: The study highlights the importance of efficient data flow, particularly in the financial sector, and suggests DQM for maintaining data quality. DJBC's import DQM level is assessed using Loshin's measurements, revealing areas for improvement through key DMBOK activities. Recommendations include data governance, strategic planning, and sequential DQM implementation. The study concludes by formulating a practical approach to be applied in a three-year span with ten indicators to measure success. Keywords: Data Quality Management, Data Quality Maturity Model, Data Quality Strategy, Loshin, DMBOK

Background: In 2020, the World Health Organization (WHO) estimated that 466 million people worldwide are affected by hearing loss, with 34 million of them being children. Indonesia is identified as one of the four Asian countries with a high prevalence of hearing loss, specifically at 4.6%. Previous research was conducted to identify diseases related to the Ear, Nose, and Throat, utilizing the certainty factor method with a test accuracy rate of 46.54%. The novelty of this research lies in the combination of two methods, the use of genetic algorithms for optimization and learning vector quantization to improve the level of accuracy for early identification of Ear, Nose, and Throat diseases. Objective: This research aims to produce a hybrid model between the genetic algorithm and the learning vector quantization neural network to be able to identify Ear, Nose, and Throat diseases with mild symptoms to improve accuracy. Methods: Implementing a 90:10 ratio means that 90% (186 data) of the data from the initial sequence is assigned for training purposes, while the remaining 10% (21 data) is allocated for testing. The procedural stages of genetic algorithm-learning vector quantization are population initialization, crossover, mutation, evaluation, selection elitism, and learning vector quantization training. Results The optimum hybrid genetic algorithm-learning vector quantization model for early identification of Ear, Nose, and Throat diseases was obtained with an accuracy of 82.12%. The parameter values with the population size 10, cr 0.9, mr 0.1, maximum epoch of 5000, error goal of 0.01, and learning rate (alpha) of 0.5. Better accuracy was obtained compared to backpropagation (64%), certainty factor 46.54%), and radial basic function (72%). Conclusion: Experiments in this research, successed identifying models by combining genetic algorithm-learning vector quantization to perform the early identification of Ear, Nose, and Throat diseases. For further research, it's very challenging to develop a model that automatically adapts the bandwidth parameters of the weighting functions during trainin Keywords: Early Identification, Ear-Nose-Throat Diseases, Genetic Algorithm, Learning Vector Quantization

Background: The field of clinical or medical imaging is beginning to experience significant advancements in recent years. Various medical imaging methods such as computed tomography (CT), X-radiation (X-ray), and magnetic resonance imaging (MRI) produce images with distinct resolution differences, goals, and noise levels, making it challenging for medical experts to diagnose diseases. Objective: The limitations of a single medical image modality have increased the necessity for medical image fusion. The proposed solution is to create a fusion method of merging two types of medical images, such as MRI and CT. Therefore, this study aimed to develop a software solution that swiftly identifies the precise region of a brain tumor, speeding up the diagnosis and treatment planning. Methods: The proposed methodology combined clinical images by using discrete wavelet transform (DWT) and inverse discrete wavelet transform (IDWT). This strategy depended on a multi-goal decay of the image information using DWT, and high-frequency sub-bands of the disintegrated images were combined using a weighted averaging method. Meanwhile, the low-frequency sub-bands were straight-forwardly replicated in the resulting image. The combined high-quality image was recreated using the IDWT. This method can handle images with various modalities and resolutions without the need for previous data. Results: The results showed that the outcomes of the proposed method were assessed by different metrics such as accuracy, recall, F1-score, and visual quality. The method showed a high accuracy of 98% over the familiar neural network techniques. Conclusion: The proposed method was found to be computationally effective and produced high-quality medical images to assist professionals. Furthermore, the method can be stretched out to other image modalities and exercised by hybrid techniques of wavelet transform and neural networks and used for different clinical image analysis tasks. Keywords: CT and MRI, Image fusion, brain tumor, wavelet transform methods, medical images, machine learning, CNN

Background: As the population increases, the need for corn products also increases. Corn is needed for various purposes, such as food consumption, industry, and animal feed. Therefore, increasing corn production is crucial to support food availability and the food industry. Objective: The objective of this project is to create a model to increase corn farming productivity using scenarios from drip irrigation systems and farmer field school programs. Methods: A system dynamics approach is utilized to model the complexity and nonlinear behaviour of the corn farming system. In addition, several scenarios are formulated to achieve the objective of increasing corn productivity. Results: Simulation results showed that adopting a drip irrigation system and operating a farmer field school program would increase corn productivity. Conclusion: The corn farming system model was successfully developed in this research. The scenario of implementing a drip irrigation system and the farmer field school program allowed farmers to increase corn productivity. Through the scenario of implementing a drip irrigation system, farmers can save water use, thereby reducing the impact of drought. Meanwhile, the scenario of the farmer field school program enables farmers to manage agriculture effectively. This study suggests that further research could consider the byproducts of corn production to increase the profits of corn farmers. Keywords: Corn Farming, Decision Support System, Modeling, Simulation, System Dynamics

Background: The rapid growth and increasing complexity of software applications are causing challenges in maintaining software quality within constraints of time and resources. This challenge led to the emergence of a new field of study known as Software Defect Prediction (SDP), which focuses on predicting future defect in advance, thereby reducing costs and improving productivity in software industry. Objective: This study aimed to address data distribution disparities when applying transfer learning in multi-project scenarios, and to mitigate performance issues resulting from data scarcity in SDP. Methods: The proposed approach, namely Transfer Learning based Low Shot Classifier (TLLSC), combined transfer learning and low shot learning approaches to create an SDP model. This model was designed for application in both new projects and those with minimal historical defect data. Results: Experiments were conducted using standard datasets from projects within the National Aeronautics and Space Administration (NASA) and Software Research Laboratory (SOFTLAB) repository. TLLSC showed an average increase in F1-Measure of 31.22%, 27.66%, and 27.54% for project AR3, AR4, and AR5, respectively. These results surpassed those from Transfer Component Analysis (TCA+), Canonical Correlation Analysis (CCA+), and Kernel Canonical Correlation Analysis plus (KCCA+). Conclusion: The results of the comparison between TLLSC and state-of-the-art algorithms, namely TCA+, CCA+, and KCCA+ from the existing literature consistently showed that TLLSC performed better in terms of F1-Measure. Keywords: Just-in-time, Defect Prediction, Deep Learning, Transfer Learning, Low Shot Learning

Background: Fraud in financial transaction is at the root of corruption issues recorded in organization. Detecting fraud practices has become increasingly complex and challenging. As a result, auditors require precise analytical tools for fraud detection. Grouping financial transaction data using K-Means Clustering algorithm can enhance the efficiency of applying Benford Law for optimal fraud detection. Objective: This study aimed to introduce Multiple Benford Law Model for the analysis of data to show potential concealed fraud in the audited organization financial transaction. The data was categorized into low, medium, and high transaction values using K-Means Clustering algorithm. Subsequently, it was reanalyzed through Multiple Benford Law Model in a specialized fraud analysis tool. Methods: In this study, the experimental procedures of Multiple Benford Law Model designed for public sector organizations were applied. The analysis of suspected fraud generated by the toolkit was compared with the actual conditions reported in audit report. The financial transaction dataset was prepared and grouped into three distinct clusters using the Euclidean distance equation. Data in these clusters was analyzed using Benford Law, comparing the frequency of the first digit’s occurrence to the expected frequency based on Benford Law. Significant deviations exceeding ±5% were considered potential areas for further scrutiny in audit. Furthermore, the analysis were validated by cross-referencing the result with the findings presented in the authorized audit organization report. Results: Multiple Benford Law Model developed was incorporated into an audit toolkit to automated calculations based on Benford Law. Furthermore, the datasets were categorized using K-Means Clustering algorithm into three clusters representing low, medium, and high-value transaction data. Results from the application of Benford Law showed a 40.00% potential for fraud detection. However, when using Multiple Benford Law Model and dividing the data into three clusters, fraud detection accuracy increased to 93.33%. The comparative results in audit report indicated a 75.00% consistency with the actual events or facts discovered. Conclusion: The use of Multiple Benford Law Model in audit toolkit substantially improved the accuracy of detecting potential fraud in financial transaction. Validation through audit report showed the conformity between the identified fraud practices and the detected financial transaction. Keywords: Fraud Detection, Benford’s Law, K-Means Clustering, Audit Toolkit, Fraudulent Practices.

Background: Due to the increased interest in cryptocurrencies, opinions on cryptocurrency-related topics are shared on news and social media. The enormous amount of sentiment data that is frequently released makes data processing and analytics on such important issues more challenging. In addition, the present sentiment models in the cryptocurrency domain are primarily focused on English with minimal work on Malay language, further complicating problems. Objective: The performance of the sentiment regression model to forecast sentiment scores for Malay news and tweets is examined in this study. Methods: Malay news headlines and tweets on Bitcoin and Ethereum are used as the input. A hybrid Generalized Autoregressive Pretraining for Language Understanding (XLNet) language model in combination with Bidirectional-Gated Recurrent Unit (Bi-GRU) deep learning model is applied in the proposed sentiment regression implementation. The effectiveness of the proposed sentiment regression model is also investigated using the multi-head self-attention mechanism. Then, a comparison analysis using Bidirectional Encoder Representations from Transformers (BERT) is carried out. Results: The experimental results demonstrate that the number of attention heads is vital in improving the XLNet-GRU sentiment model performance. There are slight improvements of 0.03 in the adjusted R2 values with an average MAE of 0.163 (Malay news) and 0.174 (Malay tweets). In addition, an average RMSE of 0.267 and 0.255 were obtained respectively for Malay news and tweets, which show that the proposed XLNet-GRU sentiment model outperforms the BERT sentiment model with lesser prediction errors. Conclusion: The proposed model contributes to predicting sentiment on cryptocurrency. Moreover, this study also introduced two carefully curated Malay corpora, CryptoSentiNews-Malay and CryptoSentiTweets-Malay, which are extracted from news and tweets, respectively. Further works to enhance Malay news and tweets corpora on cryptocurrency-related issues will be expended with implementing the proposed XLNet Bi-GRU deep learning model for greater financial insight. Keywords: Cryptocurrency, Deep learning model, Malay text, Sentiment analysis, Sentiment regression model

Background: Software engineering are courses comprising various project types, including simple assignments completed in supervised settings and more complex tasks undertaken independently by students, without the oversight of a constant teacher or lab assistant. The imperative need arises for a comprehensive assessment framework to validate the fulfillment of learning objectives and facilitate the measurement of student outcomes, particularly in computer science and software engineering. This leads to the delineation of an appropriate assessment structure and pattern. Objective: This study aimed to acquire the expertise required for assessing student performance in computer science and software engineering courses. Methods: A comprehensive literature review spanning from 2012 to October 2021 was conducted, resulting in the identification of 20 papers addressing the assessment framework in software engineering and computer science courses. Specific inclusion and exclusion criteria were meticulously applied in two rounds of assessment to identify the most pertinent studies for this investigation. Results: The results showed multiple methods for assessing software engineering and computer science courses, including the Assessment Matrix, Automatic Assessment, CDIO, Cooperative Thinking, formative and summative assessment, Game, Generative Learning Robot, NIMSAD, SECAT, Self-assessment and Peer-assessment, SonarQube Tools, WRENCH, and SEP-CyLE. Conclusion: The evaluation framework for software engineering and computer science courses required further refinement, ultimately leading to the selection of the most suitable technique, known as learning framework. Keywords: Computer science course, Software engineering course, Student assessment, Systematic literature review

Background: The primary role of a Business Intelligence (BI) system is to provide information to decision-makers within an organization. Moreover, it is crucial to acknowledge that the quality of this information is of greatest significance. Several studies have extensively discussed the importance of information quality in information systems, including BI. However, there is relatively little discussion on the factors influencing 'Information quality”. Objective: This study aimed to address this literature gap by investigating the determinants of BI maturity that impacted information quality. Methods: A maturity model comprising three dimensions was introduced, namely Data quality, BI infrastructure, and Data-driven culture. Data were collected from 84 companies and were analyzed using the SEM-PLS approach. Results: The analysis showed that maturity had a highly positive influence on Information Quality, validating the relevance of the three proposed determinant factors. Conclusion: This study suggested and strongly supported the importance and relevance of Data quality, BI infrastructure, and Data-driven culture as key dimensions of BI maturity. The robust statistical relationship between maturity and information quality showed the effectiveness of approaching the systems from a maturity perspective. This investigation paved the way for exploring additional dimensions that impact Information quality. Keywords: BI infrastructure, BI maturity, Data-driven culture, Data quality, Information quality.

Background: Artificial intelligence (AI) has become increasingly prevalent in various industries, including IT governance. By integrating AI into the governance environment, organizations can benefit from the consolidation of frameworks and best practices. However, the adoption of AI across different stages of the governance process is unevenly distributed. Objective: The primary objective of this study is to perform a systematic literature review on applying artificial intelligence (AI) in IT governance processes, explicitly focusing on the Deming cycle. This study overlooks the specific details of the AI methods used in the various stages of IT governance processes. Methods: The search approach acquires relevant papers from Elsevier, Emerald, Google Scholar, Springer, and IEEE Xplore. The obtained results were then filtered using predefined inclusion and exclusion criteria to ensure the selection of relevant studies. Results: The search yielded 359 papers. Following our inclusion and exclusion criteria, we pinpointed 42 primary studies that discuss how AI is implemented in every domain of IT Governance related to the Deming cycle. Conclusion: We found that AI implementation is more dominant in the plan, do, and check stages of the Deming cycle, with a particular emphasis on domains such as risk management, strategy alignment, and performance measurement since most AI applications are not able to perform well in different contexts as well as the other usage driven by its unique capabilities. Keywords: Artificial Intelligence, Deming cycle, Governance, IT Governance domain, Systematic literature review

Background: The learning assessment of elementary schools has recently incorporated Bloom's Taxonomy, a structure in education that categorizes different levels of cognitive learning and thinking skills, as a fundamental framework. This assessment now includes High Order Thinking Skill (HOTS) questions, with a specific focus on Indonesian topics. The implementation of this system has been observed to require teachers to manually categorize or classify questions, and this process typically requires more time and resources. To address the associated difficulty, automated categorization and classification are required to streamline the process. However, despite various research efforts in questions classification, there is still room for improvement in terms of performance, particularly in precision and accuracy. Numerous investigations have explored the use of Deep Learning Natural Language Processing models such as BERT for classification, and IndoBERT is one such pre-trained model for text analysis. Objective: This research aims to build classification system that is capable of classifying Indonesian exam questions in multiple-choice form based on Bloom's Taxonomy using IndoBERT pre-trained model. Methods: The methodology used includes hyperparameter fine-tuning, which was carried out to identify the optimal model performance. This performance was subsequently evaluated based on accuracy, F1 Score, Precision, Recall, and the time required for the training and validation of the model. Results: The proposed Fine Tuned IndoBERT Model showed that the accuracy rate was 97%, 97% F1 Score, 97% Recall, and 98% Precision with an average training time per epoch of 1.55 seconds and an average validation time per epoch of 0.38 seconds. Conclusion: Fine Tuned IndoBERT model was observed to have a relatively high classification performance, and based on this observation, the system was considered capable of classifying Indonesian exam questions at the elementary school level. Keywords: IndoBERT, Fine Tuning, Indonesian Exam Question, Model Classifier, Natural Language Processing, Bloom’s Taxonomy

Background: Multi-output Time series forecasting is a complex problem that requires handling interdependencies and interactions between variables. Traditional statistical approaches and machine learning techniques often struggle to predict such scenarios accurately. Advanced techniques and model reconstruction are necessary to improve forecasting accuracy in complex scenarios. Objective: This study proposed an Encoder-Decoder network to address multi-output time series forecasting challenges by simultaneously predicting each output. This objective is to investigate the capabilities of the Encoder-Decoder architecture in handling multi-output time series forecasting tasks. Methods: This proposed model utilizes a 1-Dimensional Convolution Neural Network with Bidirectional Long Short-Term Memory, specifically in the encoder part. The encoder extracts time series features, incorporating a residual connection to produce a context representation used by the decoder. The decoder employs multiple unidirectional LSTM modules and Linear transformation layers to generate the outputs each time step. Each module is responsible for specific output and shares information and context along the outputs and steps. Results: The result demonstrates that the proposed model achieves lower error rates, as measured by MSE, RMSE, and MAE loss metrics, for all outputs and forecasting horizons. Notably, the 6-hour horizon achieves the highest accuracy across all outputs. Furthermore, the proposed model exhibits robustness in single-output forecast and transfer learning, showing adaptability to different tasks and datasets. Conclusion: The experiment findings highlight the successful multi-output forecasting capabilities of the proposed model in time series data, with consistently low error rates (MSE, RMSE, MAE). Surprisingly, the model also performs well in single-output forecasts, demonstrating its versatility. Therefore, the proposed model effectively various time series forecasting tasks, showing promise for practical applications. Keywords: Bidirectional Long Short-Term Memory, Convolutional Neural Network, Encoder-Decoder Networks, Multi-output forecasting, Multi-step forecasting, Time-series forecasting

Background: The healthcare sector has been facing multilateral challenges regarding the quality of services and access to healthcare innovations. As the population grows, the sector requires faster and more reliable services, but the opposite is true in developing countries. As a robust technology, cloud computing has numerous features and benefits that are still to be explored. The intervention of the latest technologies in healthcare is crucial to shifting toward next-generation healthcare systems. In developing countries like Ethiopia, cloud features are still far from being systematically explored to design smart and green healthcare services. Objective: To excavate contextualized research gaps in the existing studies towards smart and green features of cloud computing in healthcare information services. Methods: We conducted a systematic review of research publications indexed in Scopus, Web of Science, IEEE Xplore, PubMed, and ProQuest. 52 research articles were screened based on significant selection criteria and systematically reviewed. Extensive efforts have been made to rigorously review recent, contemporary, and relevant research articles. Results: This study presented a summary of parameters, proposed solutions from the reviewed articles, and identified research gaps. These identified research gaps are related to security and privacy concerns, data repository standardization, data shareability, self-health data access control, service collaboration, energy efficiency/greenness, consolidation of health data repositories, carbon footprint, and performance evaluation. Conclusion: The paper consolidated research gaps from multiple research investigations into a single paper, allowing researchers to develop innovative solutions for improving healthcare services. Based on a rigorous analysis of the literature, the existing systems overlooked green computing features and were highly vulnerable to security violations. Several studies reveal that security and privacy threats have been seriously hampering the exponential growth of cloud computing. 54 percent of the reviewed articles focused on security and privacy concerns. Keywords: Cloud computing, Consolidation, Green computing, Green features, Healthcare services, Systematic literature review.

Background: The architecture of Software Defined Networking (SDN) integrated with Vehicular Ad-hoc Networks (VANETs) is considered a practical method for handling large-scale, dynamic, heterogeneous vehicular networks, since it offers flexibility, programmability, scalability, and a global understanding. However, the integration with VANETs introduces additional security vulnerabilities due to the deployment of a logically centralized control mechanism. These security attacks are classified as internal and external based on the nature of the attacker. The method adopted in this work facilitated the detection of internal position falsification attacks. Objective: This study aimed to investigate the performance of k-NN, SVM, Naïve Bayes, Logistic Regression, and Random Forest machine learning (ML) algorithms in detecting position falsification attacks using the Vehicular Reference Misbehavior (VeReMi) dataset. It also aimed to conduct a comparative analysis of two ensemble classification models, namely voting and stacking for final decision-making. These ensemble classification methods used the ML algorithms cooperatively to achieve improved classification. Methods: The simulations and evaluations were conducted using the Python programming language. VeReMi dataset was selected since it was an application-specific dataset for VANETs environment. Performance evaluation metrics, such as accuracy, precision, recall, F-measure, and prediction time were also used in the comparative studies. Results: This experimental study showed that Random Forest ML algorithm provided the best performance in detecting attacks among the ML algorithms. Voting and stacking were both used to enhance classification accuracy and reduce time required to identify an attack through predictions generated by k-NN, SVM, Naïve Bayes, Logistic Regression, and Random Forest classifiers. Conclusion: In terms of attack detection accuracy, both methods (voting and stacking) achieved the same level of accuracy as Random Forest. However, the detection of attack using stacking could be achieved in roughly less than half the time required by voting ensemble. Keywords: Machine learning methods, Majority voting ensemble, SDN-based VANETs, Security attacks, Stacking ensemble classifiers, VANETs,

Background: As a direct result of the Internet's expansion, the quantity of information shared by Internet users across its numerous platforms has increased. Sentiment analysis functions at a higher level when there are more available perspectives and opinions. However, the lack of labeled data significantly complicates sentiment analysis utilizing Bangla natural language processing (NLP). In recent years, nevertheless, due to the development of more effective deep learning models, Bangla sentiment analysis has improved significantly. Objective: This article presents a curated dataset for Bangla e-commerce sentiment analysis obtained solely from the "Daraz" platform. We aim to conduct sentiment analysis in Bangla for binary and understudied multiclass classification tasks. Methods: Transfer learning (LSTM, GRU) and Transformers (Bangla-BERT) approaches are compared for their effectiveness on our dataset. To enhance the overall performance of the models, we fine-tuned them. Results: The accuracy of Bangla-BERT was highest for both binary and multiclass sentiment classification tasks, with 94.5% accuracy for binary classification and 88.78% accuracy for multiclass sentiment classification. Conclusion: Our proposed method performs noticeably better classifying multiclass sentiments in Bangla than previous deep learning techniques. Keywords: Bangla-BERT, Deep Learning, E-commerce, NLP, Sentiment Analysis

Background: One of the latest issues in predicting cardiovascular disease is the limited performance of current risk prediction models. Although several models have been developed, they often fail to identify a significant proportion of individuals who go on to develop the disease. This highlights the need for more accurate and personalized prediction models. Objective: This study aims to investigate the effectiveness of the Grey Wolf Levenberg Model and Neural Networks in predicting cardiovascular diseases. The objective is to identify a synergistic approach that can improve the accuracy of predictions. Through this research, the authors seek to contribute to the development of better tools for early detection and prevention of cardiovascular diseases. Methods: The study used a quantitative approach to develop and validate the GWLM_NARX model for predicting cardiovascular disease risk. The approach involved collecting and analyzing a large dataset of clinical and demographic variables. The performance of the model was then evaluated using various metrics such as accuracy, sensitivity, and specificity. Results: the study found that the GWLM_NARX model has shown promising results in predicting cardiovascular disease. The model was found to outperform other conventional methods, with an accuracy of over 90%. The synergistic approach of Grey Wolf Levenberg Model and Neural Networks has proved to be effective in predicting cardiovascular disease with high accuracy. Conclusion: The use of the Grey Wolf Levenberg-Marquardt Neural Network Autoregressive model (GWLM-NARX) in conjunction with traditional learning algorithms, as well as advanced machine learning tools, resulted in a more accurate and effective prediction model for cardiovascular disease. The study demonstrates the potential of machine learning techniques to improve diagnosis and treatment of heart disorders. However, further research is needed to improve the scalability and accuracy of these prediction systems, given the complexity of the data associated with cardiac illness. Keywords: Cardiovascular data, Clinical data., Decision tree, GWLM-NARX, Linear model functions

Background: COVID-19 is a highly contagious respiratory disease with multiple mutant variants, an asymptotic nature in patients, and with potential to stay undetected in common tests, which makes it deadlier, more transmissible, and harder to detect. Regardless of variants, the COVID-19 infection shows several observable anomalies in the computed tomography (CT) scans of the lungs, even in the early stages of infection. A quick and reliable way of detecting COVID-19 is essential to manage the growing transmission of COVID-19 and save lives. Objective: This study focuses on developing a deep learning model that can be used as an auxiliary decision system to detect COVID-19 from chest CT-scan images quickly and effectively. Methods: In this research, we propose a MobileNet-based transfer learning model to detect COVID-19 in CT-scan images. To test the performance of our proposed model, we collect three publicly available COVID-19 CT-scan datasets and prepare another dataset by combining the collected datasets. We also implement a mobile application using the model trained on the combined dataset, which can be used as an auxiliary decision system for COVID-19 screening in real life. Results: Our proposed model achieves a promising accuracy of 96.14% on the combined dataset and accuracy of 98.75%, 98.54%, and 97.84% respectively in detecting COVID-19 samples on the collected datasets. It also outperforms other transfer learning models while having lower memory consumption, ensuring the best performance in both normal and low-powered, resource-constrained devices. Conclusion: We believe, the promising performance of our proposed method will facilitate its use as an auxiliary decision system to detect COVID-19 patients quickly and reliably. This will allow authorities to take immediate measures to limit COVID-19 transmission to prevent further casualties as well as accelerate the screening for COVID-19 while reducing the workload of medical personnel. Keywords: Auxiliary Decision System, COVID-19, CT Scan, Deep Learning, MobileNet, Transfer Learning

Background: Agile methodology is known for delivering effective projects with added value within a shorter timeframe, especially in Business Intelligence (BI) system which is a valuable tool for informed decision-making. However, identifying impactful elements for successful BI implementation is complex due to the wide range of Agile attributes. Objective: This research aims to systematically review and analyze the integration of BI within Agile methodology, providing valuable guidance for future projects implementation, enhancing the understanding of effective application, and identifying influential factors. Methods: Based on the Kitchenham method, 19 papers were analyzed from 288 papers, sourced from databases like Scopus, ACM, IEEE, and others published in 2016-2022. Meanwhile the extracted key factors impacting agile BI implementation were validated by qualified expert. Results: Agile was discovered to provide numerous benefits to BI projects by promoting flexibility, collaboration, and rapid iteration for enhanced adaptability, while effectively addressing challenges including those related to technology, management, and skills gaps. In addition, Agile methods, including tasks such as calculating cycle time, measuring defect backlogs, mapping code ownership, and engaging end users, offered practical solutions. The advantages included adaptability, success, value enhancement, cost reduction, shortened timelines, and improved precision. The research additionally considered other critical Agile elements such as BI tools, Agile Practices, Manifesto, and Methods, thereby enhancing insights for successful implementation. Conclusion: In conclusion, the research outlined Agile BI implementation into seven key factor groups, validated by qualified expert, providing guidance for BI integration and practices, and establishing a fundamental baseline for future applications. Keywords: Agile Methodology, Business Intelligence (BI), Expert Judgement, Kitchenham, Systematic Literature Review (SLR)

Background: The food security policy is an effort to ensure stable food availability and stable access of the community to food. As the population increases, this will affect the fulfillment of food needs in the future. Therefore, increase in rice production is needed to support food security. Objective: Conduct an analysis of the factors affecting the quality of rice production by using a dynamic system simulation that can be used as a basis for formulating policy strategies. Method: Simulation using System Dynamics (SD) is a method used to study and analyze complex systems by modeling non-linear behavior. Then several scenarios were carried out for the best decision-making using a computer. Result: The results of the scenario show that increasing the quality of paddy production in order to meet food needs in the future is doable by boosting the rendement of paddy as it will upgrade rice production which will contribute greatly to rice production. Conclusion: From the simulation results, the study can be used to increase the quality of rice production to maintain food security by improving the harvesting mechanism to increase yields. For further research, the use of Smart Agriculture can be considered to increase production of rice. Keywords: Food security, Rice production, Rice production, System dynamics

Background: Gamification is a trend that has emerged with the growth of e-commerce. Given the wide range of human characteristics, determining which gamification elements perform well and what impact those gamification elements have can be challenging. Objective: This study aims to conduct a systematic literature review to broadly review the impact that can be caused by the application of gamification elements in e-commerce. This study also attempts to identify the current trends in using gamification elements. Methods: This study was carried out based on the Kitchenham approach and analyzes 25 research papers extracted from a total of 550 papers. The articles were gathered from ACM, Emerald, ScienceDirect, and Scopus and were published between 2016 and 2021. Results: This study found that the trend of research in the field of gamification in e-commerce continues to grow every year. Also, this study found that the most frequently used gamification elements are achievement-oriented (such as rewards, points, badges, and leaderboards). Meanwhile, immersion-related gamification elements (such as avatars, fantasy, etc.) are emerging as a new trend for new gamification elements to be incorporated in e-commerce. This study also found three major themes, namely consumer loyalty, consumer engagement, and user behavior, as a result of the application of gamification in e-commerce. Conclusion: This study helps to improve knowledge of various gamification elements, trends, and impacts on e-commerce. Future studies need to examine the challenges that may arise in the application of gamification elements to the three major themes found in this study and find potential solutions to overcome them. Keywords: E-Commerce, Gamification, Gamification trends and applications, Kitchenham, Systematic literature review.

Background: The concept of customer orientation, which is based on a set of fundamental beliefs that prioritize the interests of the customer, requires companies to detect these interests in order to maintain a high level of quality in their products or services. Furthermore, there are several indicators of customer orientation, and one of them is their opinion or taste, which provides valuable feedback for businesses. With the rapid development of social media, customers can express emotions, thoughts, and opinions about services or products that may not be easily conveyed in the real world. Objective: The objective of this study is to detect customer orientation towards product or service quality, as expressed in online or social media. Additionally, the study showcases the novelty and superiority of the annotation process used for detecting customer orientation classifications. Methods: This study employs a method to compare the classification performance of the Vader lexicon annotation process with manual annotation. To accomplish this, a dataset from the Amazon website will be analyzed and classified using the Support Vector Machine algorithm. The objective of this method is to determine the level of customer orientation present within the dataset. To evaluate the effectiveness of the Vader lexicon, the study will compare the results of manual and automatic data annotation. Results: The results showed that customer orientation towards product or service quality has a predominantly positive value, comprising up to 76% of the total responses analyzed. Conclusion: The findings demonstrate that using Vader in the annotation process results in superior accuracy values compared to manual annotation. Specifically, the accuracy value increased from 86% to 88.57%, indicating that Vader could be a reliable tool for annotating text. Therefore, future studies should consider using Vader as a classifier or integrating it into the annotation process to further enhance its performance. Keywords: Classification, Customer, Orientation, Text analysis, Vader lexicon,

Background: Software testing and software security have become one of the most important parts of an application. Many studies have explored each of these topics but there is a gap wherein the relation of software security and software testing in general has not been explored. Objective: This study aims to conduct a systematic literature review to capture the current state-of-the-art in software testing related to security. Methods: The search strategy obtains relevant papers from IEEE Xplore and ScienceDirect. The results of the search are filtered by applying inclusion and exclusion criteria. Results: The search results identified 50 papers. After applying the inclusion/exclusion criteria, we identified 15 primary studies that discuss software security and software testing. We found approaches, aspects, references, and domains that are used in software security and software testing. Conclusion: We found certain approach, aspect, references, and domain are used more often in software security testing Keywords: Software security, Software testing, Security testing approach, Security threats, Systematic literature review

Background: Feature extraction process is noteworthy in order to categorize brain tumors. Handcrafted feature extraction process consists of profound limitations. Similarly, without appropriate classifier, the promising improved results can’t be obtained. Objective: This paper proposes a hybrid model for classifying brain tumors more accurately and rapidly is a preferable choice for aggravating tasks. The main objective of this research is to classify brain tumors through Deep Convolutional Neural Network (DCNN) and Support Vector Machine (SVM)-based hybrid model. Methods: The MRI images are firstly preprocessed to improve the feature extraction process through the following steps: resize, effective noise reduction, and contrast enhancement. Noise reduction is done by anisotropic diffusion filter, and contrast enhancement is done by adaptive histogram equalization. Secondly, the implementation of augmentation enhances the data number and data variety. Thirdly, custom deep CNN is constructed for meaningful deep feature extraction. Finally, the superior machine learning classifier SVM is integrated for classification tasks. After that, this proposed hybrid model is compared with transfer learning models: AlexNet, GoogLeNet, and VGG16. Results: The proposed method uses the ‘Figshare’ dataset and obtains 96.0% accuracy, 98.0% specificity, and 95.71% sensitivity, higher than other transfer learning models. Also, the proposed model takes less time than others. Conclusion: The effectiveness of the proposed deep CNN-SVM model divulges by the performance, which manifests that it extracts features automatically without overfitting problems and improves the classification performance for hybrid structure, and is less time-consuming. Keywords: Adaptive histogram equalization, Anisotropic diffusion filter, Deep CNN, E-health, Machine learning, SVM, Transfer learning.

Background: During the Covid-19 period, the government made policies dealing with it. Policies issued by the government invited public opinion as a form of public reaction to these policies. The easiest way to find out the public’s response is through Twitter’s social media. However, Twitter data have limitations. There is a mix between facts and personal opinions. It is necessary to distinguish between these. Opinions expressed by the public can be both positive and negative, so correlation is needed to link opinions and their emotions. Objective: This study discusses sentiment and emotion detection to understand public opinion accurately. Sentiment and emotion are analyzed using Pearson correlation to determine the correlation. Methods: The datasets were about public opinion of Covid-19 retrieved from Twitter. The data were annotated into sentiment and emotion using Pearson correlation. After the annotation process, the data were preprocessed. Afterward, single model classification was carried out using machine learning methods (Support Vector Machine, Random Forest, Naïve Bayes) and deep learning method (Bidirectional Encoder Representation from Transformers). The classification process was focused on accuracy and F1-score evaluation. Results: There were three scenarios for determining sentiment and emotion, namely the factor of aspect-based and correlation-based, without those factors, and aspect-based sentiment only. The scenario using the two aforementioned factors obtained an accuracy value of 97%, while an accuracy of 96% was acquired without them. Conclusion: The use of aspect and correlation with Pearson correlation has helped better understand public opinion regarding sentiment and emotion more accurately. Keywords: Aspect-based sentiment, Deep learning, Emotion detection, Machine learning, Pearson correlation, Public opinion.

Background: COVID-19 is a disease that attacks the respiratory system and is highly contagious, so cases of the spread of COVID-19 are increasing every day. The increase in COVID-19 cases cannot be predicted accurately, resulting in a shortage of services, facilities and medical personnel. This number will always increase if the community is not vigilant and actively reduces the rate of adding confirmed cases. Therefore, public awareness and vigilance need to be increased by presenting information on predictions of confirmed cases, recovered cases, and cases of death of COVID-19 so that it can be used as a reference for the government in taking and establishing a policy to overcome the spread of COVID-19. Objective: This research predicts COVID-19 in confirmed cases, recovered cases, and death cases in Lampung Province Method: This study uses the ANN method to determine the best network architecture for predicting confirmed cases, recovered cases, and deaths from COVID-19 using the k-fold cross-validation method to measure predictive model performance. Results: The method used has a good predictive ability with an accuracy value of 98.22% for confirmed cases, 98.08% for cured cases, and 99.05% for death cases. Conclusion: The ANN method with k-fold cross-validation to predict confirmed cases, recovered cases, and COVID-19 deaths in Lampung Province decreased from October 27, 2021, to January 24, 2022. Keywords: Artificial Intelligence, Artificial Neural Network (ANN) K-Fold Cross Validation, COVID-19 Cases, Data Mining, Prediction.

Background: COVID-19 has become a primary public health issue in various countries across the world. The main difficulty in managing outbreaks of infectious diseases is due to the difference in geographical, demographic, economic inequalities and people's behavior in each region. The spread of disease acts like a series of diverse regional outbreaks; each part has its disease transmission pattern. Objective: This study aims to assess the association of socioeconomic and demographic factors to COVID-19 cases through cluster analysis and forecast the daily cases of COVID-19 in each cluster using a predictive modeling technique. Methods: This study applies a hierarchical clustering approach to group regencies and cities based on their socioeconomic and demographic similarities. After that, a time-series forecasting model, Facebook Prophet, is developed in each cluster to assess the transmissibility risk of COVID-19 over a short period of time. Results: A high incidence of COVID-19 was found in clusters with better socioeconomic conditions and densely populated. The Prophet model forecasted the daily cases of COVID-19 in each cluster, with Mean Absolute Percentage Error (MAPE) of 0.0869; 0.1513; and 0.1040, respectively, for cluster 1, cluster 2, and cluster 3. Conclusion: Socioeconomic and demographic factors were associated with different COVID-19 waves in a region. From the study, we found that considering socioeconomic and demographic factors to forecast COVID-19 cases played a crucial role in determining the risk in that area. Keywords: COVID-19, Facebook Prophet , Hierarchical clustering, Socioeconomic and demographic

Background: The utilization of virtual healthcare services, particularly telemedicine, has been accelerated by the COVID-19 pandemic. Although the pandemic is no longer the primary concern, telemedicine still holds potential for long-term adoption. However, implementing telemedicine in Indonesia as an online platform for remote healthcare delivery still faces issues, despite its potential. Further investigation is required to identify the factors that affect its adoption and develop strategies to surmount implementation challenges. Objective: This study aims to examine and enrich knowledge about the adoption of telemedicine in Indonesia. Methods: A cross-sectional survey was conducted through an online questionnaire to collect data. Unified Theory of Acceptance and Use of Technology 2 (UTAUT2) was employed by integrating with several factors, such as eHealth Literacy, Privacy Concerns, and Trust. Gender and age were considered as moderating variables. Data samples were analyzed using Partial Least Square – Structural Equation Modeling (PLS–SEM). Results: The findings suggest that performance expectancy, effort expectancy, social influence, eHealth literacy, and trust have a significant impact on adults’ behavioral intention to use telemedicine. However, facilitating condition, price value, and privacy concern do not show any significant effects on adults’ Behavioral Intention to Use Telemedicine. Conclusion: This study highlights the importance of understanding adoption factors to develop effective strategies. Results show performance expectancy, effort expectancy, social influence, eHealth literacy, and trust are significant factors, while facilitating condition, price value, and privacy concern are not. The UTAUT2 model is a good predictive tool for healthcare adoption. To increase usage intention, several aspects must be considered in the implementation of telemedicine. Keywords: Adoption, Behavioral Intention to Use, Telemedicine, UTAUT2, Virtual Healthcare.

Background: Twitter is a popular social media where users express emotions, thoughts, and opinions that cannot be channelled in the real world. They do this by tweeting short, concise, and clear messages. Since users often express themselves, Twitter data can detect mental health trends. Objective: This study aims to detect suicidal messages through tweets written by users with mental health issues. Methods: These tweets are analysed and classified using the lexicon-based and Naive Bayes algorithms to determine whether it contains suicidal messages. Results: The classification results show that the ‘normal’ classification is predominant at 52.3% of the total 3,034,826 tweets, which indicates an increase from September to December 2021. Conclusion: Most tweets are categorised as ‘normal’, therefore the mental health status appears secure. However, this finding needs to be re-examined in the future, especially in DKI Jakarta Province, which has the most cases of mental disorders. This study found that the Naive Bayes algorithm is more accurate (85.5%) than the lexicon-based algorithm. This can be improved in future studies by increasing performance at the pre-processing stage. Keywords: Lexicon Based, Mental Disorder, Mental Health, Naïve Bayes, Twitter

Background: Dental disease detection is essential for the diagnosis of dental diseases. Objective: This research compares the Mamdani fuzzy logic and Naïve Bayes in detecting dental diseases. Methods: The first is to process data on dental disease symptoms and dental support tissues based on complaints of toothache consulted with experts at a community health centre (puskesmas). The second is to apply the Mamdani fuzzy logic and the Naïve Bayes to the proposed expert system. The third is to provide recommended decisions about dental diseases based on the symptom data inputted into the expert system. Patient data were collected at the North Cilacap puskesmas between July and December 2021. Results: The Mamdani fuzzy logic converts uncertain values into definite values, and the Naïve Bayes method classifies the type of dental disease by calculating the weight of patients’ answers. The methods were tested on 67 patients with dental disease complaints. The accuracy rate of the Mamdani fuzzy logic was 85.1%, and the Naïve Bayes method was 82.1%. Conclusion: The prediction accuracy was compared to the expert diagnoses to determine whether the Mamdani fuzzy logic method is better than the Naïve Bayes method. Keywords: Dental Disease, Expert System, Mamdani Fuzzy Logic, Naïve Bayes, Prediction

Background: The rapid development of telecommunication technology has prompted the creation of various messenger applications. The competition among social messengers to gain market share is becoming tighter. Objective: This study aims to capture user preferences for messenger platforms and inform software development companies to improve their products based on user needs. Methods: This research uses quantitative methods, i.e., categorical analysis and multiple linear regression analysis, to extend the results from qualitative methods that identify the preferences in past studies. The data were obtained through a questionnaire. Results: The results show that customers are influenced by accessibility, flexibility, effectiveness and chat history. Meanwhile, users are influenced by responsiveness, user-friendly interface, performance, personal needs, privacy and security, and customer services. Conclusion: The research can identify the indicators to guide the creation of an ideal messenger platform based on customer and user preferences. Keywords: Conjoint, Messenger Platform, Multiple Linear Regression, Preference

Background: Radiologists used chest radiographs to detect coronavirus disease 2019 (COVID-19) in patients and determine the severity levels. The COVID-19 cases were grouped into five classes, each receiving different treatments. An intelligent system is needed to advance the detection and identify vector features of X-ray images with a quality that is too poor to be read by radiologists. Deep learning is an intelligent system that can be used in this case. Objective: The current study compares the classification and accuracy of detection methods with two, three dan five classes. Methods: Deep learning can classify visual geometry group VGG 19 architectures with 1000 classes. The classification of the five classes' convolutional neural network (CNN) underwent model validation with a confusion matrix to produce accuracy and class values. The system could then diagnose patients’ examinations by radiology specialists. Results: The results of the five-class method showed 98% accuracy, the three-class method showed 99.99%, and the two-class showed 99.99%. Conclusion: It can be concluded that using the VGG 19 model is effective. This system can classify and diagnose viruses in patients to assist radiologists by reading the images. Keywords: COVID-19, CNN, Classification, Deep Learning

Background: Understanding customers’ electricity consumption patterns is essential for developing predictive analytics, which is needed for effective supply and demand management. Objective: This study aims to understand customers’ segmentation and consumption behaviour using a hybrid approach combining the K-Means clustering, customer lifetime value concept, and analytic hierarchy process. Methods: This study uses more than 16 million records of customers’ electricity consumption data from January 2019 to December 2020. The K-Means clustering identifies the initial market segments. The results were then evaluated and validated using the customer lifetime value concept and analytical hierarchy process. Results: Three customer segments were identified. Segment 1 has 282 business customers with a total capacity of 938,837 kWh, peak load usage of 27,827 kWh, and non-peak load usage of 115,194 kWh. Segment 2 has 508,615 business customers with a total capacity of 4,260 kWh, a peak load of 35 kWh, and a non-peak load of 544 kWh. Segment 3 has 37 business customers with a total capacity of 2,226,351 kWh, a peak load of 123.297 kWh, and a non-peak load of 390,803. Conclusion: A business strategy that could be taken is to base customer relationship management (CRM) on the three-customer segmentation. For the least profitable segment, aside from retail account marketing, a continuous partnership program is needed to increase electricity consumption during the non-peak period. For the highly and moderately profitable segments, a premium business-to-business approach can be applied to accommodate their increasing energy consumption without excessive electricity use in the peak period. Special account executives need to be deployed to handle these customers.