JICO: International Journal of Informatics and Computing

Custom Concept Text-to-Image Using Stable Diffusion Model in Generative Artificial Intelligence

Alam Rahmatulloh — Thu, 30 May 2024 00:00:00 +0000

The ability of algorithms to produce content that closely mimics human work has revolutionized several fields thanks to generative artificial intelligence, or Gen AI. However, these developments also raise questions about generative models' transparency, predictability, and behavior. Considering the relevance of this topic and the expanding influence of AI on society, research into it is imperative. This paper aims to empirically explore the nuances of behavior in the setting of discriminative generative AI, using the stable diffusion model as an example. We will be better equipped to handle obstacles and guarantee the ethical and responsible application of generative AI in a world that is changing quickly if we have a deeper grasp of this phenomenon. The research method is carried out in several stages, such as dataset collection, modeling, testing, and analysis of results. The research results show that generative artificial intelligence can create realistic images like the original. However, there are still several challenges, including the availability of a reasonably large dataset for training data and high and long computing times. Likewise, the results of the Fréchet Inception Distance (FID) test were still quite large, namely 1284.4430, which shows that the quality of this model is still not good.

Sentiment Analysis of Application X on The Impact of Social Media Content on Adolescent Mental Well-Being using Naïve Bayes Algorithm

Randi Rizal, Ulka Chandini Pendit, Nuraminah Ramli, Siti Annisa — Thu, 30 May 2024 00:00:00 +0000

Since the pandemic, the use of social media has increased significantly. However, its presence has raised significant concerns about its impact on the mental well-being of teenagers. The pervasive influence of social media has led to substantial changes in the social system within society. Despite this influence, there is currently no comprehensive understanding of the specific impact of social media on mental health. To address this gap, this research proposes the use of sentiment analysis of social media posts with the Naive Bayes algorithm as an approach to identify and classify positive and negative sentiments in these posts related to the mental well-being of teenagers. This solution aims to provide a deeper understanding of the impact of social media content on this vulnerable demographic. In this study, a total of 555,361 social media posts were successfully collected and analyzed using the Naive Bayes algorithm, which was trained with a sample of 27,977 test data. The research results demonstrate that sentiment analysis with the Naive Bayes algorithm is effective in classifying social media sentiment, with 50.55% of the posts classified as positive and 46.97% classified as negative. The identified sentiment patterns have provided valuable insights into the positive and negative impact of social media content on the mental well-being of teenagers.

Advanced Phishing Attack Detection Through Network Forensic Methods and Incident Response Planning Based on Machine Learning

Siti Rahayu Selamat, Randi Rizal, Cucu Nursihab, Nashihun Amien — Thu, 30 May 2024 00:00:00 +0000

The widespread use of smartphones has led to an increase in cybercrimes, particularly phishing attacks. Phishing attacks are commonly propagated through email, WhatsApp groups, and other communication channels. The stolen data is then used to commit further crimes, exploiting the victims' personal information. This study addresses the detection of phishing attacks using network forensic methods and incident response planning. Unlike previous approaches that relied solely on Incident Response Plans (IRPs) and Incident Handling methods to react to phishing attacks, this research emphasizes proactive detection. By employing network forensics, suspicious websites can be identified and differentiated from legitimate ones, enabling early detection and prevention of phishing attacks. The results demonstrate that network forensics can significantly enhance the ability to detect phishing sites before they can harm users. In our experiments, we analyzed a dataset of 10,000 websites, identifying 95% of phishing sites with a false positive rate of only 2%. Utilizing the Random Forest machine learning algorithm, we achieved high performance metrics with an accuracy of 96.5%, precision of 97.1%, recall of 95.8%, and an F1-score of 96.4%. This proactive approach not only mitigates the risk of phishing but also provides a robust framework for incident response, ensuring that potential threats are identified and neutralized promptly.

Optimizing Data Management in Web Applications through Google Drive API Integration and Synchronization

Putri Septia Amalia, Erna Haerani, Rusnida Romli, Trisna Ari Roshinta — Thu, 30 May 2024 00:00:00 +0000

The rise of Web-based applications has created a demand for streamlined data management and automatic data synchronization. Even manually stored local data is often insufficient to meet these requirements, necessitating a solution that can efficiently manage data access and storage through Cloud technology. This study advocates for utilizing the Google Drive API to resolve these issues. By leveraging the benefits of Google Drive's Cloud storage, Web applications can seamlessly synchronize user-uploaded data to the Cloud. To initiate this integration, a Google account is required to authenticate the process and serve as a mediator for data exchange. This approach employs secure authentication and authorization mechanisms to ensure data privacy. The system is developed using an iteration-based approach starting with user requirements analysis, followed by interface design and API integration. Pilot tests were then conducted to validate system performance under various usage scenarios. The findings revealed a noteworthy advancement in the synchronization and administration of data through the Web-based application with a data transmission duration of under 60 seconds, contingent on internet speed. Google Drive's API integration enables users to access files and manage them in real-time, surpassing prior limitations. To meet the demands of progressively intricate Web-based applications, future research can concentrate on enhancing data security and optimizing performance.

SMOTE Technique Utilization in Cirrhosis Classification: A Comparison of Gradient Boosting and XGBoost

Abdul Latip, Huzain Azis, Hidayatulloh Himawan, Dian Ade Kurnia — Thu, 30 May 2024 00:00:00 +0000

Cirrhosis is a chronic liver disease with significant health implications, responsible for 56,585 deaths annually, and ranking as the 9th leading cause of mortality worldwide. Early detection is crucial for effective treatment and better patient outcomes, as cirrhosis can progress to irreversible damage if not addressed in its initial stages. This research focuses on developing an advanced, integrated method for detecting cirrhosis by employing a combination of Synthetic Minority Over-sampling Technique (SMOTE) and machine learning models, specifically Gradient Boosting and XGBoost. The use of SMOTE is critical in this study as it addresses class imbalance in the dataset, which is a common challenge in medical diagnosis problems, especially when dealing with rare or minority conditions like cirrhosis. Class imbalance can lead to biased models that perform poorly on the minority class, which, in this case, could mean missing crucial cirrhosis diagnoses. SMOTE oversamples the minority class to ensure a more balanced dataset, which improves the model's ability to detect cirrhosis accurately. The research further includes a performance comparison between two powerful machine learning algorithms: Gradient Boosting and XGBoost. Gradient Boosting is known for its ability to optimize the model by focusing on misclassified instances in a sequential manner, while XGBoost, an advanced version of Gradient Boosting, is renowned for its speed and efficiency due to parallel processing and advanced regularization techniques.

Digitograph: A Mobile Digital Signatures Application for PDF file Using ED25519 and Asymmetric Encryption

Annisa Putri Wahyuni, Arif Bramantoro, Randi Rizal, Souhayla Elmeftahi — Thu, 30 May 2024 00:00:00 +0000

Digital signatures have become an essential tool in the digital era, providing a secure and efficient way to authenticate and verify the integrity of digital documents. The increasing need for remote and electronic transactions has led to a surge in the development of digital signature technology. This research presents a mobile application, Digitograph, designed to facilitate the process of digitally signing PDF files using ED25519 and Asymmetric Encryption. Several processes were employed to complete the Digitograph application, including a literature study to gather information and documents related to the development process. A research framework was prepared to ensure that the processes in the study were carried out in a directed and systematic manner. The development of the Digitograph application was successfully accomplished, with significant results demonstrating improvements in security, efficiency, and ease of use for digital signatures on PDF files. The following are some key aspects of the Digitograph application's development: 1) Enhanced security, 2) Performance and efficiency, 3) User-friendliness. Digital signatures using ED25519, and Asymmetric Encryption are one of the key technological applications in modern cryptography. ED25519 offers a high level of security, efficiency, and ease of use. This method enhances data security and significantly simplifies key management to address vulnerabilities in Asymmetric Encryption. The Digitograph application ensures the authenticity and integrity of documents, providing a practical solution in the digital era.