International Journal of Artificial Intelligence

Development of a Student Expense Tracking System Using Optical Character Recognition

Muhammad Hairil Shaharudin, Ahmad Fadli Saad, Achmad Yani, Abdi Manaf — Sat, 28 Jun 2025 09:44:11 +0000

Personal financial literacy is a vital skill for university students, yet many struggle to track their daily expenses due to time constraints and low awareness. This study aims to design and develop a web-based Student Expense Tracking System using Optical Character Recognition (OCR) technology to address this issue. The system allows users to automatically extract and record spending information from receipt images, reducing manual input and improving financial awareness. The development followed the Web Development Life Cycle (WDLC) using the Waterfall model, comprising planning, design, development, and testing phases. Visual Studio Code, Python 3, and Tesseract OCR were employed in system implementation. Wireframes and mockups guided the interface design, while backend development focused on data storage and OCR integration. Functionality testing showed a 100% pass rate across ten scenarios, validating the system's performance in image processing, budget management, and spending visualization. Usability testing using the Post-Study System Usability Questionnaire (PSSUQ) with 30 participants yielded a mean score of 4.45 out of 5, indicating a high level of user satisfaction. The system scored highest on ease of use (4.6), visual design (4.7), and recommendation likelihood (4.8), confirming its intuitive interface and appeal. Slightly lower scores in user confidence (4.1) and data organization (4.2) point to opportunities for interface refinement and improved user guidance. This research concludes that OCR can effectively support financial tracking for students. Future enhancements with NLP and machine learning are recommended to automate expense categorization and improve analytical capabilities.

Evaluation of Perplexity and Syntactic Handling Capabilities of ClueAI Models on Japanese Medical Texts

Tatsuhiro Haga, Keiyo Matsumoto, Ippei Asahiko, Shunzo Mizoguchi — Sat, 28 Jun 2025 09:48:49 +0000

This study aims to evaluate the effectiveness of a large Japanese language model, ClueAI, tailored to the medical domain, in the task of predicting Japanese medical texts. The background of this study is the limitations of general language models, including multilingual models such as multilingual BERT, in handling linguistic complexity and specific terminology in Japanese medical texts. The research methodology includes fine-tuning the ClueAI model using the MedNLP corpus, with a MeCab-based tokenization approach through the Fugashi library. The evaluation is carried out using the perplexity metric to measure the model's generalization ability in predicting texts probabilistically. The results show that ClueAI that has been tailored to the medical domain produces lower perplexity values than the multilingual BERT baseline, and is better able to understand the context and sentence structure of medical texts. MeCab-based tokenization is proven to contribute significantly to improving prediction accuracy through more precise morphological analysis. However, the model still shows weaknesses in handling complex syntactic structures such as passive sentences and nested clauses. This study concludes that domain adaptation provides improved performance, but limitations in linguistic generalization remain a challenge. Further research is recommended to explore models that are more sensitive to syntactic structures, expand the variety of medical corpora, and apply other Japanese language models in broader medical NLP tasks such as clinical entity extraction and classification.

A Policy Analysis of the Danish National AI Strategy: Ethical and Governance Implications for AI Ecosystems

Henrik Lauritsen, David Hestbjerg, Lone Pinborg, Christensen Pisinger — Sat, 28 Jun 2025 09:52:48 +0000

The Danish National AI Strategy presents a structured approach to building an ethical and innovative AI ecosystem. It emphasizes four main pillars: ethical AI development, public data utilization, skills development, and strategic technology investment. The strategy has achieved notable success, especially in the education sector, where ethical principles like fairness, transparency, and accountability are well-integrated. However, issues such as algorithmic bias and fairness remain, indicating the need for ongoing refinement of ethical frameworks. Public data plays a central role in AI innovation, particularly in healthcare and education. Yet, challenges related to data privacy and access continue to pose obstacles, highlighting the importance of robust data governance. Skills development programs have helped prepare the workforce for AI-related roles, though limited employer participation, especially among small businesses, suggests the need for more inclusive outreach. Furthermore, while government and private funding have supported advanced AI research, the transition from innovation to practical application still faces gaps. This study employed a qualitative descriptive approach, utilizing document analysis and thematic analysis based on data from government publications and expert interviews with 20 stakeholders, including policymakers and AI specialists. The findings provide valuable insights into Denmark’s AI journey and serve as a reference for other countries aiming to implement responsible, inclusive, and sustainable AI strategies.

Network Anomaly Detection System using Transformer Neural Networks and Clustering Techniques

Sat, 28 Jun 2025 09:54:54 +0000

This study proposes a hybrid approach for network anomaly detection by integrating a Transformer-based model with clustering techniques. The methodology begins with the application of K-means clustering as a preprocessing step to group similar network traffic data, thereby reducing data complexity and highlighting significant patterns. The clustered data is then fed into a Transformer model, which utilizes multi-head self-attention mechanisms to capture intricate temporal dependencies and contextual relationships within sequential data. This dual-stage approach enhances the model’s ability to differentiate between normal and anomalous behaviors in network traffic. Trained on a network security dataset, the system effectively identifies both common and rare attack types. According to the results, the suggested ensemble classifier outperformed existing deep learning models with an accuracy of over 99.5%, 98.5%, and 99.9% on the UNSW-NB15 dataset. The synergy between the unsupervised pattern recognition of clustering and the deep learning capabilities of Transformers enables a scalable and adaptable solution for real-world network security applications, making it suitable for proactive cyber threat detection and mitigation.

CAD for Robotics: Trends, Opportunities, Considerations, and Constraints

Sat, 28 Jun 2025 09:58:32 +0000

In Southeast Asia, the integration of Computer-Aided Design (CAD) into robotics development has become increasingly vital in meeting the growing demand for rapid, task-specific automation. CAD plays a central role in enhancing how robotic systems are designed, simulated, and prototyped—enabling improved design precision, reduced development time, and accelerated innovation. This paper investigates the current trends in CAD applications for robotics and highlights key opportunities, including collaborative design approaches, rapid prototyping capabilities, and the convergence of digital engineering tools. Furthermore, the study discusses critical technical considerations such as software interoperability, real-time simulation integration, and the need for upskilling in CAD-related competencies. Drawing from both academic research and industrial practice across Southeast Asian countries, the findings reveal a pressing need for tighter integration between CAD platforms, robotic simulation environments, and control systems. The analysis identifies several regional challenges, including limited access to advanced CAD tools, inconsistent adoption in educational curricula, and disparities in technical training infrastructure. The paper concludes with strategic recommendations to support the growth of CAD-driven robotics in the region: bridging the digital skills gap, improving access to design technologies, promoting cross-institutional collaboration, and encouraging targeted research to adapt CAD tools to local industrial needs. These efforts are crucial for enabling Southeast Asia to capitalize on CAD’s transformative potential in developing agile, affordable, and application-specific robotic solutions.