JOIV : International Journal on Informatics Visualization
Not a member yet
786 research outputs found
Sort by
A Model for Classification Usability Testing Practically from the Agile Methodology Aspect
Full text linkUsability is a crucial feature in the success of software products. Development practitioners know that to preserve and improve the quality of the product and usability has to be carefully considered throughout the development process. The tendency toward empowering users in software development drives the ongoing pursuit of solutions to reconcile agile and usability goals. In this paper, the authors aim to develop a model for classifying usability testing methods according to aspects of agile methodologies. This model can enable agile practitioners to obtain end-user feedback when implementing usability tests at the appropriate time and place during development and thus produce useful and usable software. Mixed methods (qualitative and quantitative) were used in this research to collect primary and secondary data. This research adopted the convenience non-probability sampling technique for evaluating the model. The evaluation determines whether it could provide valuable information supporting consistent usability tests. The method of performance profiles is also applied in this evaluation to gain accurate results and avoid any biases that might unnecessarily affect the outcomes. The evaluation results were encouraging, and the model showed beneficial effects in integrating usability work into an agile approach, especially since all attributes showed high importance among participants' accepted satisfaction, representing the least essential scale. The developed model must be applied practically to the other integration models in future work. Furthermore, several observation techniques are required to thoroughly cover the integration by software development teams from diverse organizations.
A Review of AMQP Protocol: Characteristics, Security Challenges and Proposed Enhancement
Full text linkFor distributed networks, especially in Internet of Things (IoT) applications, the Advanced Message Queuing Protocol (AMQP) is a prevalent messaging protocol. It provides a versatile messaging system for asynchronous communication between applications or components in distributed networks. Applications for the Internet of Things (IoT) commonly use machine-to-machine (M2M) communication protocols, including CoAP, AMQP, XMPP, and MQTT. To better understand AMQP and its use in IoT contexts, this paper aims to examine its features, security issues, and possible improvements. Using a literature review methodology, the study examines relevant studies published between 2019 and 2024. The results demonstrate that AMQP offers key benefits for IoT communication, including reliable communication, message integrity, authentication, and encryption. Given these characteristics, data transmission between devices is more secure, making AMQP a good option for various IoT applications. Nevertheless, the analysis also highlights essential security issues, including weaknesses that could be exploited against sensitive industries such as smart cities and healthcare. To overcome these issues, the paper proposes improvements to AMQP, including the addition of Time-To-Live (TTL) features and negative acknowledgment mechanisms. The implications for future studies include a more thorough examination of AMQPs and other IoT protocols' security and privacy features, as well as the development of robust security measures to protect IoT ecosystems from potential cyber threats. The study enhances the understanding of AMQP's role in IoT communication and underscores the need for further research to strengthen its security and effectiveness in complex IoT scenarios
Artificial Intelligence Adoption on Investment Platform for Robo Advisory Users in Indonesia
Full text linkRobo-advisors provide an alternative financial solution tailored for regular clients. Beyond the acceptability of technology, financial factors significantly influence the adoption of robo-advisors. While existing studies extensively discuss the stages involved in the intention to utilize robo-advisors, only a few offer insights into financial capabilities. The purpose of this study is to investigate the extent to which Indonesian investors embrace robo-advisors by incorporating financial variables such as financial goals and financial literacy into the technology adoption in Robo Advisor. Additionally, the study explores the relationship between application costs and data privacy on the adoption of robo-advisor technology. This research employs a quantitative approach using purposive sampling techniques. Data were collected through a survey of 431 robo-advisor users and analyzed using SmartPLS. The findings reveal a significant and positive correlation between financial goals, perceived technology usefulness, and application costs in the adoption of robo-advisors in Indonesia. These results contribute to the development of investment decision theory using technology-based approaches, specifically robo-advisors. Furthermore, companies in the financial sector, particularly in wealth management or investment management, can benefit from incorporating financial goal features, enhancing technological performance, and setting competitive fees to increase adoption rates. Future research should further explore robo-advisor adoption, focusing on additional financial variables and financial behaviors that drive technology adoption as an investment decision. These findings highlight the importance of considering both financial and technological factors in promoting the use of robo-advisors among investors especially in Indonesia
Recommendation System for Mobile-Based Oil Palm Fertilization Period with Rainfall Prediction using ANN
Full text linkWeather conditions significantly affect human activities, including the oil palm plantation sector, which in practice considers weather and climate conditions. Oil palm is an annual crop that requires proper nutrition throughout the year. Plant nutrition through fertilization must be according to the specific needs of oil palms. Knowing the type of fertilizer, calculating the dosage, and evaluating the climatic characteristics significantly affect the effectiveness and efficiency of fertilization. According to one palm oil farmer, fertilization should ideally be done when the soil is moist or not during the dry season so plants can absorb fertilizers properly. If fertilization is ineffective, then the operational costs of plant maintenance to buy fertilizers become less efficient. Due to climate change, farmers often find it difficult to determine the optimal timing of fertilization. Therefore, rainfall prediction is essential. Thus, fertilization can run well and get maximum results. The recommendation system in this research includes a rainfall prediction system with machine learning methods and an Artificial Neural Network. The recommendation system is a mobile-based application that allows oil palm farmers to obtain information on the appropriate time to fertilize based on rainfall. The evaluation of rainfall prediction using ANN has the MSE value of 0.0019981 and the MAPE value of 9.355%. It can be concluded that the rainfall prediction model is working optimally. This system can be combined with harvesting forecasting and recommendations of oil palm plantation periods to become a monitoring system for oil palm productivity
Agent-Oriented Modelling and Simulation for Robotic Based Predator Control
Full text linkAgent-oriented modeling (AOM) is a methodology that can develop complex tasks that involve multi-agent system development, such as robotic systems. There are still insufficient studies on how Agent-Oriented Modelling benefits robotic development. There is little reference to using Agent-Oriented Modelling to develop complex systems, especially robotic applications. This study aims to investigate the adoption of AOM for robotic surveillance modeling and simulation for predator control in the farming sector and to conduct qualitative comparisons on robotic models and simulation methods. A case study of robotic-based predator control is used to develop the system using the AOM model. Later, this is incorporated into a Netlogo simulation to illustrate the suggested methodology in the model simulation stage. A qualitative analysis of the model is performed to validate the model. SUS analysis for AOM usability at the score of 68.35 shows AOM is at the average usability level for beginner users in software development. Qualitative analysis shows that beginner users prefer to use AOM for complex adaptive and distributed robotic systems. AOM is introduced to create robotic-based predator control in a structured manner to prove that AOM can be used to develop complex systems. Introducing Agent-Oriented Modelling in various domains leads to higher confidence in the industry player's adoption of this model across multiple system developments
Systematic Literature Review of Gender Bias within Video Games Character Design
Full text linkGender bias in video games refers to the unequal treatment and discrimination that players experience based on gender, which is often normalized within the gaming community. Gender bias is widespread in Multiplayer Online Battle Arena (MOBA) games, where it can take many different forms. Common examples include assumptions made about players' abilities, character design in games, and the roles given to characters according to gender. This situation has created an unwelcoming environment, especially for female players, leading to feelings of exclusion. This study conducts a systematic literature review to examine gender bias in MOBA games, explicitly focusing on character representation, hypersexualized character models, and gameplay mechanics. By analyzing data from peer-reviewed articles, theses, and research papers, the study highlights the recurring patterns of bias and identifies gaps in current approaches. Although prior studies have explored the elements that contribute to gender bias, few studies have offered practical solutions to mitigate this bias. However, there is still a lack of research proposing a practical game design framework that integrates strategies to reduce this bias. In conclusion, efforts to address gender bias are not only significant in terms of design ethics, but also a good strategy in expanding the game's audience. This study identifies possible solutions that might help future research and be developed into a conceptual framework model that developers can understand to create a more inclusive, fair, and profitable gaming environment in the long term
Development of a Decision Support System Based on New Approach Respond to Criteria Weighting Method and Grey Relational Analysis: Case Study of Employee Recruitment Selection
Full text linkThe purpose of this research is to propose a new approach in the criteria weighting method using the RECA method, the RECA method can help provide a systematic and structured framework for determining criteria weights in multi-criteria decision making. The determination of weights using the RECA method is to increase objectivity and accuracy in the candidate assessment and selection process by determining the appropriate weight for each criterion based on responses and assessments from experts or stakeholders. Testing the RECA Method with Multi Attribute Decision Making (MADM) techniques is an important step in measuring the effectiveness of the RECA Method in the context of multi-criteria decision making. Ranking tests using Spearman correlation between the RECA method and other methods such as SAW with a correlation value of 1, MOORA with a correlation value of 0.9636, MAUT with a correlation value of 0.9515, WP with a correlation value of 0.891, SMART with a correlation value of 0.9636, and TOPSIS with a correlation value of 0.8788 show a high level of rank consistency between the RECA method and these methods. This indicates that the RECA Method has a strong ability to generate similar candidate rankings with other methods, validating its reliability and consistency in the context of multi-criteria decision making. Implications for further research include exploring the application of the RECA method in different decision-making contexts other than recruitment, such as performance evaluation, project selection, or supplier selection. Further research could investigate the integration of the RECA method with other decision-making methods or algorithms to improve its performance and applicability in complex decision environments. Comparative studies with larger sample sizes and diverse datasets can provide deeper insights into the effectiveness and reliability of the RECA method compared to other methods
Chest X-Ray Images Clustering Using Convolutional Autoencoder for Lung Disease Detection
Full text linkIn healthcare, medical imaging is commonly used for health assessments. One of the most commonly used types of medical imaging is X-ray imaging. One area that often undergoes examination using this modality is the lungs, where healthcare professionals use X-ray images to interpret the results. However, prolonged interpretation of X-ray results by healthcare professionals and other work activities can lead to errors and potentially result in invalid disease identification. There is a need for a system that can classify the detection results from these images to assist healthcare professionals in their tasks. Various methods can be used for this purpose, such as classification, clustering, segmentation, etc. However, data labeling requires significant resources and costs, especially with large-scale datasets. One possible solution is to use an unsupervised learning approach to address this. One method under unsupervised learning is clustering, which allows the system to process and understand data patterns without needing external annotations or manual labeling. This research uses an autoencoder as a subcategory of unsupervised learning. This is because autoencoders can automatically extract relevant features from the data without needing external label guidance. The research utilizes a dataset consisting of 700 X-ray images of the chest, including 500 images showing disease and 200 normal X-ray images. This research aims to determine the effectiveness of clustering methods using an autoencoder model in grouping X-ray image results. The research conducted two experiments. In the first experiment, an autoencoder with 18 Layers was used, resulting in the best performance with a value of K=15 and a rand index of 76%. In the second experiment, an autoencoder with a reduced number of Layers (11 Layers) was used, and it achieved the best performance with a value of K=15 and a rand index of 87%
Adaptive Deep Convolution Neural Network for Early Diagnosis of Autism through Combining Personal Characteristic with Eye Tracking Path Imaging
Full text linkAutism is a large set of illnesses related to brain development, also referred to as autism spectrum disorder (ASD). According to WHO reports, 1 in 100 children is expected to have ASD. Numerous behavioral domains are affected, including linguistic, interpersonal skills, stereotypical and repetitive behaviors which represent an extreme instance of a neurodevelopmental abnormality. Identifying ASD can be difficult and exhausting because its symptoms are remarkably identical to those of many other disorders of the mind. Medical professionals can improve diagnosis efficiency by adapting deep learning practices. In clinics for autism spectrum disorders, eye-tracking scan pathways (ETSP) have become a more common instrument. This approach uses quantitative eye movement analysis to study attentional processes, and it exhibits promising results in the development of indicators that can be used in clinical studies for autism. ASD can be identified by comparing the abnormal attention span patterns of children’s having the disorder to the children’s who are typically developing. The recommended model makes use of two publicly viable datasets, namely ABIDE and ETSP imaging. The proposed deep convolutional network consists of four hidden convolution layers and uses 5-fold cross-validation strategy. The performance of the proposed model is validated against multilayer perceptron (MLP) and conventional machine learning classifiers like decision tree (DT), k-nearest neighbor (KNN) and Random Forest (RF) using metrics like sensitivity, specificity and area under curve (AUC). The findings demonstrated that without the need for human assistance, the suggested model is capable of correctly identifying children with ASD
Multi-Objective k-Nearest Neighbor for Breast Cancer Detection
Full text linkEarly detection of cancer is crucial. This study aims to increase the efficiency of breast cancer detection using the modified k-nearest neighbor (k-NN) algorithm. Since k-NN faces challenges with sensitivity to k values and computational complexity, a modification of k-NN was proposed, namely a multi-objective k-NN model. It was developed to incorporate multi-objective optimization and local density to create a more robust and efficient classification algorithm. The model dynamically determines the k value based on the sample density, optimizing accuracy and efficiency. Breast cancer data were collected from the University of Wisconsin Hospitals, Madison. The experimental results showed that the multi-objective k-NN model outperformed traditional k-NN and k-NN with feedback support. The proposed model achieved an accuracy of 93.7%, with precision values of 93% for the negative cancer class and 94% for the positive cancer class. These results indicate that the multi-objective k-NN model provides superior accuracy and precision in breast cancer detection, demonstrating its potential for clinical applications