cover
Contact Name
Hariyadi Fajar Nugroho
Contact Email
aisa@journals.cognispectra.com
Phone
+6285725769193
Journal Mail Official
aisa@journals.cognispectra.com
Editorial Address
Sejahtera Street Number 15, Gumpang, Kartasura, Sukoharjo, Indonesia
Location
Kab. sukoharjo,
Jawa tengah
INDONESIA
Artificial Intelligence Systems and Its Applications (AISA)
ISSN : -     EISSN : 31100457     DOI : https://doi.org/10.65917/aisa.v1i2.2025
Core Subject : Science,
Artificial Intelligence Systems and Its Applications (AISA) is an international, peer-reviewed journal publishing cutting-edge original research in Artificial Intelligence (AI) and its applications. The journal explores theory, methodologies, and real-world applications of AI in various domains, including but not limited to machine learning, natural language processing, AI-driven embedded systems, AI-integrated solutions, and computational social science. AISA aims to serve both academic researchers and industry practitioners by providing an effective platform for the timely dissemination of advanced AI innovations and emerging trends. The journal welcomes contributions that address fundamental challenges in AI, interdisciplinary approaches, and critical applications of AI across different fields. Scope AISA invites high-quality submissions in the following areas: Artificial Intelligence and Its Applications – Core advancements and breakthroughs in AI technologies. Machine Learning and Its Applications – Algorithms, models, and learning paradigms. Natural Language Processing (NLP) – Language understanding, text generation, and conversational AI. AI Embedded Systems – AI in IoT, robotics, and smart hardware solutions. AI Integrated Systems – AI-powered automation, decision-making, and intelligent computing. Computational Social Science – AI applications in social media analytics, sentiment analysis, and human behavior modeling. Publication Information Peer-Reviewed: Ensuring rigorous evaluation and high-quality contributions. Open Access: Providing unrestricted access to cutting-edge AI research. Frequency: Published quarterly with special issues on emerging AI trends. AISA welcomes original research papers, review articles, and case studies that contribute to the advancement and practical implementation of AI technologies. The journal aims to bridge the gap between theoretical AI advancements and their real-world applications, fostering innovation in artificial intelligence systems globally.
Articles 10 Documents
Sentiment Analysis of Public Feedback on Universitas Muhammadiyah Surakarta Through Google Maps Reviews: Insights and Implications Agus Ardiansyah Nh; Endang Wahyu Pamungkas; Sohail Akhtar
Artificial Intelligence Systems and Its Applications Vol. 1 No. 1 (2025): Vol. 1, No. 1, June 2025
Publisher : CV Cognispectra Publishing

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.65917/aisa.v1i1.12

Abstract

In the era of digital feedback, higher education institutions face growing challenges in making sense of large volumes of user-generated reviews, especially those written in multiple languages. This study analyzes 1,717 Google Maps reviews related to Universitas Muhammadiyah Surakarta (UMS), collected over five years in Bahasa Indonesia and English. To overcome limitations of manual and monolingual sentiment analysis, we employed a pre-trained multilingual transformer model—lxyuan/distilbert-base-multilingual-cased-sentiments-student—without additional fine-tuning. The analysis revealed that 88% of reviews were classified as positive, with most praise directed at campus facilities, while criticism often targeted administrative services. Beyond sentiment classification, this study explored text length, confidence scores, and user engagement patterns to uncover deeper behavioral insights. We also developed SentiMu, an interactive dashboard that visualizes sentiment trends, recent reviews, word clouds, and key metrics, enabling university stakeholders to monitor feedback in real time. The dashboard was built using Next.js and FastAPI for optimal performance and scalability. By automating the analysis and visualization of multilingual online reviews, this study provides a practical and scalable framework for institutions to understand student and visitor experiences, supporting data-driven decisions to enhance campus services and reputation.
Utilizing Artificial Intelligence for Personalized Digital Learning Nendy Nendy Akbar Rozaq Rais
Artificial Intelligence Systems and Its Applications Vol. 1 No. 1 (2025): Vol. 1, No. 1, June 2025
Publisher : CV Cognispectra Publishing

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.65917/aisa.v1i1.13

Abstract

The digital transformation in education has opened up significant opportunities for leveraging artificial intelligence (AI) to personalize learning. This study aims to explore the use of AI in adapting content, methods, and learning pace to individual learner characteristics. The research method employed is a literature study with a descriptive qualitative approach, involving the analysis of various scientific sources from the last five years relevant to AI and digital education. The findings indicate that AI can enhance learning motivation, accelerate concept comprehension, and improve academic outcomes through features such as content recommendation systems, personal tutors, and adaptive assessments. Furthermore, AI supports more interactive and emotional learning experiences through the use of learning agents such as chatbots. Nevertheless, the implementation of AI faces several challenges, including data privacy, teacher readiness, infrastructure limitations, and algorithmic bias. The discussion highlights the need for regulation, educator training, and the development of ethical and inclusive systems to optimize the benefits of AI. These findings demonstrate that AI has great potential to revolutionize the digital learning ecosystem, but its implementation must be conducted wisely and responsibly.
AI Customer Service in the Digital Business World: Understanding Customer’s Thoughts on Ai-Driven Personalization in E-Commerce and Social Media Ariel Siffrin; Raaziq Akbar Al Qarni
Artificial Intelligence Systems and Its Applications Vol. 1 No. 1 (2025): Vol. 1, No. 1, June 2025
Publisher : CV Cognispectra Publishing

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.65917/aisa.v1i1.14

Abstract

AI in the digital business world means the use of artificial intelligence technology to improve efficiency, productivity, and customer experience in various operational and strategic areas of a company. The Focus of this study is to see customer thoughts on the implementation of artificial intelligence in the online shopping sector. This study will be carried out in 4 different phases, including; comprised of data searching, data inputting, data processing, and editing phase. This research data was processed using the SmartPLS (Partial Least Squares) application.  The study's findings indicate that artificial intelligence (AI) has profoundly transformed the manner in which e-commerce platforms engage with consumers. The results of our study demonstrate that the provision of effective AI services can significantly improve customer satisfaction. These findings underscore the importance of implementing AI in e-commerce marketing strategies, as by understanding customer preferences and providing relevant recommendations, companies can increase customer engagement and loyalty. This study underscores the significance of AI services in enhancing customer experience and paves the way for further research that can deepen our understanding of the interaction between technology and consumer behavior in the context of e-commerce
Analysis of AI Algorithm Development: From Machine Learning to Deep Learning Jompon Pitaksantayothin; Hariyadi Fajar Nugroho
Artificial Intelligence Systems and Its Applications Vol. 1 No. 1 (2025): Vol. 1, No. 1, June 2025
Publisher : CV Cognispectra Publishing

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.65917/aisa.v1i1.18

Abstract

The development of Artificial Intelligence (AI) is currently very rapid, but there is still much confusion regarding the differences and evolution of the main algorithms, namely machine learning (ML) and deep learning (DL). This study aims to analyze the development of AI algorithms conceptually and technically from conventional ML to DL, and to provide a structured understanding of the paradigm shift in AI development. The method used is a systematic literature study of 10 recent scientific articles discussing aspects of ML and DL algorithms. The results of the analysis show that ML relies on manual feature extraction with the advantages of computational efficiency and interpretability, while DL is able to process large and complex data automatically with better performance, although it requires high computing resources and faces interpretability challenges. The discussion also identifies the main challenges that AI still faces as well as innovation opportunities to overcome these limitations. In conclusion, a deep understanding of the evolution of AI algorithms is essential as a foundation for the development of more adaptive, effective, and transparent AI technology in the future.
How Well Do Vision-Language Models Explain Sarcasm? An Evaluation of Multimodal Explanation Quality for Social Media Posts Ikhlasul Amal; Annisa Nur Ramadhani
Artificial Intelligence Systems and Its Applications Vol. 1 No. 1 (2025): Vol. 1, No. 1, June 2025
Publisher : CV Cognispectra Publishing

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.65917/aisa.v1i1.22

Abstract

Sarcasm is a complex communicative phenomenon frequently encountered in social media, where the literal meaning of language sharply contradicts the speaker’s true intent, often reinforced by multimodal cues such as incongruent images or memes. While prior research has primarily focused on detecting sarcasm, far less attention has been devoted to generating human-interpretable explanations that clarify why content is sarcastic. This study addresses this gap by systematically evaluating the capabilities of fifteen Vision–Language Models (VLMs) of varying parameter sizes to produce multimodal sarcasm explanations under zero-shot and few-shot learning conditions. Using the publicly available MORE dataset of social media posts annotated with concise human-written explanations, we benchmarked each model’s outputs with three widely used evaluation metrics, including ROUGE, BERTScore, and Sentence-BERT, to assess both surface-level overlap and deeper semantic alignment. Our findings reveal that smaller models can rival or even outperform larger architectures in n-gram similarity measures, while embedding-based metrics often yield high scores even when generated explanations contradict the ground truth. These results highlight the limitations of current automatic metrics in reliably capturing the nuanced reasoning underlying sarcasm. Overall, this work demonstrates that model scale does not consistently predict explanation quality and underscores the need for more robust evaluation protocols.
Text Representation Method Analysis and Its Implementation in Automatic Essay Scoring System Alya Zakhira Anjani; Divi Galih Prasetyo Putri; Widhy Hayuhardhika Nugraha Putra
Artificial Intelligence Systems and Its Applications Vol. 1 No. 2 (2025): Vol. 1, No. 2, December 2025
Publisher : CV Cognispectra Publishing

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.65917/aisa.v1i2.30

Abstract

The automatic essay scoring system is one of many problems in terms of natural language processing (NLP) that has long been studied. This study used an approach using text similarity with cosine similarity method to determine correct and incorrect predictions in an automatic essay scoring system. However, the text representation phase is also an important phase. This study compares the performance of three text representation methods in their implementation into an automatic essay scoring system. The methods are Indonesian Version of Bidirectional Encoder from Transformers (IndoBERT), Embeddings from Language Model (ELMo), and FastText. In addition, the combination of each method with WordNet as an additional lexical resource is also compared. The result of comparison using dataset “Indonesian Query Answering Dataset for Online Essay Test System” shows that the combination of IndoBERT and WordNet model has the best performance proven with highest accuracy achieved being 0.69, precision being 0.54, recall being 0.81, and F1-score being 0.48. Then the model was implemented as an essay evaluation feature development for the Certified Government Accounting Associate (CGAA) Exam Simulation site. The feature performance test results show an average load time of 418.8 ms when accessed by 10 users simultaneously and 15064 ms when accessed by 100 users simultaneously. The features developed are expected to be able to support the evaluation process more efficiently.
Integrating SHAP Guided Feature Optimization into Gradient Boosting for Explainable Machine Learning Muhammad Hikmal Yazid
Artificial Intelligence Systems and Its Applications Vol. 1 No. 2 (2025): Vol. 1, No. 2, December 2025
Publisher : CV Cognispectra Publishing

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.65917/aisa.v1i2.34

Abstract

Artificial Intelligence (AI) has achieved remarkable success in predictive modeling, yet the lack of explainability in complex models remains a major challenge for adoption in high-stakes domains. This study addresses this problem by developing a machine learning pipeline that integrates explainability techniques with high-performance predictive models. The objectives are to enhance model transparency, evaluate performance on real-world datasets, and compare the proposed approach with conventional baseline models. Experimental evaluation was conducted on healthcare and finance datasets, using gradient boosting models combined with SHAP explanations to provide feature-level interpretability. The results demonstrate that the proposed approach achieves 92.5% accuracy, 91.2% precision, and 90.8% recall, outperforming baseline models while maintaining transparent decision-making. Visualization of feature contributions confirmed that the model’s predictions align with domain knowledge, enhancing trust and accountability. The study highlights the feasibility of balancing predictive performance with explainability, providing a practical framework for deploying AI in critical applications. Limitations include increased computational requirements for large-scale datasets. The findings offer implications for both researchers and practitioners by demonstrating that highly accurate models can remain interpretable, promoting ethical and responsible AI deployment. Future work should explore scalability, real-time interpretability, and application to additional domains, further bridging the gap between predictive power and model transparency.
Deep Learning-Based Image Recognition System for Automated Microplastic Detection and Water Pollution Monitoring Faiz Ni'matul Haq
Artificial Intelligence Systems and Its Applications Vol. 1 No. 2 (2025): Vol. 1, No. 2, December 2025
Publisher : CV Cognispectra Publishing

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.65917/aisa.v1i2.43

Abstract

Microplastic pollution in aquatic ecosystems poses a serious environmental and public health risk, requiring effective and scalable monitoring technologies. Current detection methods, which rely on manual microscopy and spectroscopic verification, are labor-intensive, time-consuming, and unsuitable for large-scale assessments. While deep learning offers a potential alternative, current approaches are often limited by dependence on non-public datasets and a lack of model interpretability. This paper presents an automated, transparent, and repeatable deep learning system for microplastic identification based on advanced YOLO (You Only Look Once) architectures. The proposed system utilizes and evaluates YOLOv8 and YOLOv11 models on a consolidated public dataset of microplastic images, using extensive data augmentation to enhance model robustness. Results show that the YOLOv11 model achieves a state-of-the-art mean Average Precision (mAP@50) of 94.7%, significantly outperforming the YOLOv8 baseline at 89.5%. Additionally, implementing Explainable AI (XAI) techniques, particularly Eigen-CAM, provides vital visual validation of the model's decision-making process by highlighting microplastic features, thereby improving interpretability and confidence. This study offers a repeatable, highly accurate, and transparent detection framework suitable for automated environmental monitoring. The findings demonstrate that Transformer-based object detection models combined with XAI can significantly enhance microplastic pollution assessment, supporting more effective monitoring and mitigation of aquatic pollution.
Application of Machine Learning with XGBoost for Classifying Chemical Compound Activity as Potential Alzheimer’s Drug Candidates Muhibbul Tibri; Rahmat Sufri; Teuku Rizky Noviandy
Artificial Intelligence Systems and Its Applications Vol. 1 No. 2 (2025): Vol. 1, No. 2, December 2025
Publisher : CV Cognispectra Publishing

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.65917/aisa.v1i2.44

Abstract

Alzheimer’s disease is a progressive neurodegenerative disorder characterized by cognitive and memory decline, with acetylcholinesterase (AChE) as one of the most important therapeutic targets. Conventional experimental screening of AChE inhibitors is time-consuming, costly, and prone to high failure rates. Therefore, computational approaches based on machine learning are increasingly adopted to accelerate early-stage drug discovery. This study aims to classify the bioactivity of chemical compounds against AChE as potential Alzheimer’s drug candidates using the Extreme Gradient Boosting (XGBoost) algorithm. Bioactivity data were obtained from the ChEMBL database, where IC50 values were converted into pIC50 and classified into active and inactive compounds. Molecular descriptors were calculated using the Mordred library, and the dataset was divided into training and testing sets with an 80:20 ratio. Hyperparameter optimization was performed using Random Search to improve model performance. The experimental results show that the baseline XGBoost model achieved an accuracy of 84.39%, while the optimized model improved accuracy to 86.90% with an AUC of 0.9343. SHAP analysis revealed that descriptors related to electronic properties and lipophilicity, such as SssCH2, PEOE_VSA7, and SlogP_VSA, contributed most significantly to compound activity classification. These findings demonstrate that XGBoost combined with explainable AI techniques is effective for in silico identification of potential Alzheimer’s drug candidates and provides meaningful insights into relevant molecular features
Multi-Objective Bio-Inspired Hyperparameter Optimization for Trustworthy Brain Tumor MRI Classification Using Calibration-Aware CNN Models Kafitra Marna Ibrahim; Zaky Zaujan Jayaputra
Artificial Intelligence Systems and Its Applications Vol. 1 No. 2 (2025): Vol. 1, No. 2, December 2025
Publisher : CV Cognispectra Publishing

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.65917/aisa.v1i2.45

Abstract

Automated brain tumor classification from magnetic resonance imaging (MRI) has become an essential component in advancing computer-aided diagnosis. However, many deep learning approaches prioritize accuracy alone while overlooking two key requirements for real-world medical deployment: the reliability of predicted confidence scores and the computational efficiency required for clinical integration. This study proposes a multi-objective bio-inspired hyperparameter optimization framework to produce convolutional neural network (CNN) models that are accurate, well-calibrated, and computationally efficient. The model is optimized using a Multi-Objective Particle Swarm Optimization (MOPSO) algorithm that jointly minimizes validation error, Expected Calibration Error (ECE), and inference latency. Experiments were conducted on a four-class Brain Tumor MRI dataset, and the optimized configuration achieved a test accuracy of 95 percent, an ECE of 1.48 percent, and a sub-millisecond inference latency of 0.88 milliseconds per sample. Grad-CAM visualizations further confirm that the model’s decisions are guided by clinically relevant tumor regions. The results demonstrate that multi-objective hyperparameter optimization offers a robust pathway for developing trustworthy, efficient, and interpretable artificial intelligence systems for medical imaging applications.

Page 1 of 1 | Total Record : 10