International Journal of Intelligent Systems最新文献

{"title":"Temporal Heterogeneous Network Representation Learning With Dynamic Influence Modeling","authors":"Haodan Ran,&nbsp;Yang Fang,&nbsp;Xiang Zhao,&nbsp;Jiuyang Tang,&nbsp;Weiming Zhang","doi":"10.1155/int/6673499","DOIUrl":"https://doi.org/10.1155/int/6673499","url":null,"abstract":"<p>Temporal heterogeneous network representation learning is a pivotal approach for encapsulating the diversity of nodes and edges along with their temporal evolution into concise, low-dimensional node representations. This technique has demonstrated remarkable efficacy in various network analysis and inference tasks. However, existing approaches study network evolution mainly by analyzing snapshots of temporal networks, while neglecting the intrinsic formation mechanisms of temporal heterogeneous networks. Few dynamic models delve into the intrinsic factors propelling network evolution. To fill this research gap, we introduce a novel learning framework for temporal heterogeneous network representation learning with dynamic influence modeling, denoted as THNRD. THNRD pioneers the application of the Hawkes process to temporal heterogeneous networks, utilizing the linking process of dynamic events to emulate the network’s formation mechanism, capturing the intrinsic dynamic progression of temporal heterogeneous networks. Subsequently, THNRD introduces a multilayer spatiotemporal aggregation model under a unified spatiotemporal framework, which is designed to harmoniously integrate the semantic and dynamic attributes of the networks. We also take node influence into consideration to further describe the temporal emergent phenomena. We verify the effectiveness of our proposed method via extensive experimental evaluations on real-world datasets. The results consistently demonstrate that THNRD outperforms current state-of-the-art methods.</p>","PeriodicalId":14089,"journal":{"name":"International Journal of Intelligent Systems","volume":"2026 1","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/int/6673499","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147668453","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Large Language Models in Cellulose Biopolymer Studies: Evaluating ChatGPT and Microsoft Copilot for Information and Reference Accuracy","authors":"Mesbah Ahmad,&nbsp;Tanmay Rahman,&nbsp;Nitesh Kumar Kasera,&nbsp;Shoeb Ahmed","doi":"10.1155/int/2260439","DOIUrl":"https://doi.org/10.1155/int/2260439","url":null,"abstract":"<p>With the increasing reliance on large language models (LLMs) for scientific research, it is critical to assess their reliability in specialized fields such as biopolymer science, particularly with respect to the verifiability of the references. This study examines the performance of two widely used LLMs, ChatGPT (GPT-4 omni) and Microsoft Copilot (GPT-4), in responding to questions and citing the references for the answers on cellulose biopolymers. The questions are set based on three cognitive levels: beginner, intermediate, and expert, and the accuracy of the responses and references provided by the models are assessed. Results show that ChatGPT outperforms Copilot in all cognitive levels, particularly in addressing mathematical problems. ChatGPT achieves 91.9%, 91.9%, and 88.6% accuracy at the beginner, intermediate, and expert levels, respectively, whereas Copilot achieves 82.4%, 82.5%, and 71.1% accuracy. However, the analysis of references reveals critical shortcomings in both models. While Copilot tends to cite more recent journal articles, ChatGPT often relies on older ones. In many cases, references are incomplete, fabricated, or lack proper context, highlighting the persistent challenge of verifying AI-generated citations. Overall, Copilot outperforms ChatGPT in providing correct references. Results show that ChatGPT provides 6%, 36%, and 45% fabricated references at the beginner, intermediate, and expert levels, respectively, whereas Copilot delivers 13%, 9%, and 7% fabricated references at these levels. This work emphasizes that while LLMs hold promise in supporting scientific inquiry in biopolymers, their current limitations in response accuracy and citation reliability need to be addressed before they can serve as dependable tools for scholarly work.</p>","PeriodicalId":14089,"journal":{"name":"International Journal of Intelligent Systems","volume":"2026 1","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/int/2260439","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147668452","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Smoke Detection Algorithm Based on Improved Feature Fusion","authors":"Xiaoyun Wu,&nbsp;Jie Zhao","doi":"10.1155/int/2269004","DOIUrl":"https://doi.org/10.1155/int/2269004","url":null,"abstract":"<p>The smoke detection algorithm has wide application value in fire warning, environmental monitoring, and other fields. To improve the accuracy and real-time performance of smoke detection, a smoke detection algorithm based on improved feature fusion is proposed. A new feature matrix is constructed by combining the Haar-like algorithm and the LBP algorithm. To improve the two feature extraction methods, the study compares the pixel values of the neighborhood and the central pixel and weights the pixels of each region according to the difference. The improved LBP matrix is obtained. The longitudinal, transverse, diagonal, and cross-shaped feature differences of the Haar template are extracted, and these feature differences are used as the weighting coefficients of the Haar feature matrix. Then, the extracted features are input into the improved YOLOv7, and the efficient pyramid split attention network (EPSANet) is introduced into YOLOv7. Meanwhile, the dual compressed attention mechanism (shuffle attention network, SANet) integrates spatial channels. EPSANet extracts multilevel smoke feature information by squeezing and allocating attention to features of different scales through a pyramid-shaped structure. SANet further enhances the model’s focus on key features and reduces background interference by integrating a dual compression attention mechanism of space and channels. The results show that the detection accuracy of the research design method in different practical application scenarios is above 85%, and the first detection time is also within 5s. The detection error value of the research method mainly fluctuates in the range of 0.02∼0.04, which has high detection accuracy and detection efficiency. In summary, it can be seen that the smoke detection method designed in this study can provide more reliable technical support for fire warning, environmental monitoring, and other fields.</p>","PeriodicalId":14089,"journal":{"name":"International Journal of Intelligent Systems","volume":"2026 1","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-04-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/int/2269004","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147668197","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Intelligent Modeling of Protein Function and Drug Discovery Using Large Language Models: A Review","authors":"Minna Zhang,&nbsp;Bin Liu,&nbsp;Jing Xu,&nbsp;Baozhen Zhou,&nbsp;Hongkai Sun,&nbsp;Yangyang Wang,&nbsp;Jihan Wang,&nbsp;Mengju Xue","doi":"10.1155/int/9139921","DOIUrl":"https://doi.org/10.1155/int/9139921","url":null,"abstract":"<p>Large language models (LLMs), originally developed for natural language processing, are increasingly being applied to biological data, offering a novel paradigm for intelligent modeling in protein function prediction and drug discovery. By leveraging their ability to capture contextual dependencies in sequences, LLMs such as ProteinBERT, ESM, and BioGPT have shown strong potential in learning informative representations from protein sequences and biomedical literature. These models can support downstream tasks including function annotation, protein–protein interaction prediction, and de novo drug design. This review presents a comprehensive overview of recent advances in applying LLMs to molecular biology, focusing on their architectures, training strategies, and integration with domain-specific knowledge. We highlight the strengths and current limitations of LLM-based approaches, including challenges in data scarcity, interpretability, and biological relevance. Finally, we discuss future research directions for enhancing the reliability, efficiency, and domain adaptation of LLMs in life sciences. This work aims to provide a foundation for researchers seeking to apply intelligent systems based on LLMs in computational biology and drug development.</p>","PeriodicalId":14089,"journal":{"name":"International Journal of Intelligent Systems","volume":"2026 1","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-04-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/int/9139921","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147667980","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Uncovering Regulatory Networks Through Residual Graph Learning of circRNA–miRNA Interactions","authors":"Murtada K. Elbashir,&nbsp;Madallah Alruwaili,&nbsp;Mahmood Mohamed","doi":"10.1155/int/7155544","DOIUrl":"https://doi.org/10.1155/int/7155544","url":null,"abstract":"<p>Circular RNAs (circRNAs) and microRNAs (miRNAs) are key regulators of gene expression, and their interactions are involved in many biological processes and diseases. Nevertheless, the experimental validation of circRNA–miRNA interactions (CMIs) remains challenging due to resource and technical constraints; therefore, robust computational models are essential for accurate CMI prediction. This paper proposes a new residual graph learning framework (RGLRE) for CMIs prediction using role-aware graph embeddings. RGLRE combines attention-based message transmission, DropEdge regularization, and similarity-based hard negative sampling. Role2Vec embeddings are employed in RGLRE to provide role-aware structural representations of circRNA and miRNA nodes, which support stable information propagation in the residual graph neural network. These design choices enable more robust and generalizable feature learning for CMI prediction. The experimental analysis of various datasets reveals that RGLRE is more effective in capturing complicated biological dependencies compared with the existing models like CMAGN, BEROLECMI, and GAT. RGLRE was evaluated on the CMI-9905 and CMI-9589 datasets with 5-fold cross-validation, and it showed better results than the current techniques. More precisely, the model showed accuracy of 0.9115 on CMI-9905 and 0.9161 on CMI-9589, and the overall score of the model consistently obtained high AUROC and AUPR values. The framework’s modular design and scalability make it a promising tool for broader biological interaction prediction tasks.</p>","PeriodicalId":14089,"journal":{"name":"International Journal of Intelligent Systems","volume":"2026 1","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/int/7155544","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147585117","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Combining Clinical Characteristics With CTA Radiomics for Predicting Intracranial Aneurysm Rupture Status","authors":"Ye Shi,&nbsp;Peipei Wang,&nbsp;Mingquan Ye,&nbsp;Aiping Wu,&nbsp;Yunfeng Zhou","doi":"10.1155/int/1564250","DOIUrl":"https://doi.org/10.1155/int/1564250","url":null,"abstract":"<p>Intracranial aneurysms (IAs) are clinically categorized as ruptured or unruptured. Size variability complicates precise segmentation and rupture assessment. This study integrates deep learning, machine learning, clinical characteristics, and computed tomography angiography (CTA) radiomics to determine IA rupture status. A dataset of 443 aneurysms (101 unruptured and 342 ruptured) was curated from affiliated hospitals. IAs were segmented via Swin UNETR, with radiomic features extracted via PyRadiomics. Following dimensionality reduction, five classifiers, support vector machine (SVM), logistic regression (LR), random forest (RF), multilayer perceptron (MLP), and voting classifier, were evaluated via the area under the receiver operating characteristic curve (AUC-ROC). Among the 1074 radiomic features, 25 were significantly correlated with rupture status. The RF, voting, MLP, LR, and SVM classifiers achieved AUCs of 0.94, 0.93, 0.88, 0.88, and 0.84, respectively. These results demonstrate the discriminative power of the combined t-test, LASSO, and PCA feature selection methods.</p>","PeriodicalId":14089,"journal":{"name":"International Journal of Intelligent Systems","volume":"2026 1","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/int/1564250","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147615445","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing Machine Learning Models for Mental Health Classification Through Iterative Training and Text-Based Augmentation","authors":"Suparna Das,&nbsp;Kamil Reza Khondakar,&nbsp;Hirak Mazumdar,&nbsp;Ajeet Kaushik,&nbsp;Sunil Kumar Singh","doi":"10.1155/int/2620320","DOIUrl":"https://doi.org/10.1155/int/2620320","url":null,"abstract":"<p>Machine learning (ML) models are frequently used to classify mental health information from textual data, but their practical use is constrained by their poor interpretability and lack of tools to fix training-related reasoning errors. Explainable AI (XAI) approaches currently in use mostly offer post hoc explanations without methodically utilizing explanation quality to enhance model performance. This paper proposes an explanation-driven iterative learning framework for classifying texts related to mental health in order to close this gap. Using local interpretable model–agnostic explanations (LIME), the suggested approach produces explanations for model predictions. These explanations are then quantitatively assessed by comparing them to ground-truth explanations using cosine similarity. The models undergo iterative retraining on the enlarged dataset after data samples linked to low-quality explanations are selectively enhanced with text generated by GPT-3.5. The framework is tested on social media–based mental health datasets using a variety of deep learning– and transformer-based models, such as LSTM, Bi-LSTM, BERT variants, and GPT-3.5. According to experimental results, transformer-based models perform better and show steady accuracy gains over time, with overall gains of roughly 4%–7%. The suggested method improves interpretability and predictive accuracy, providing a reliable and scalable solution for high-stakes NLP applications such as reliable mental health classification.</p>","PeriodicalId":14089,"journal":{"name":"International Journal of Intelligent Systems","volume":"2026 1","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-03-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/int/2620320","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147615088","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Fractional-Order Self-Learning Sparrow Search Algorithm for Multithreshold Image Segmentation","authors":"Shoutong Huang,&nbsp;Yu Ma,&nbsp;Ruiyang Wu,&nbsp;Zhen Li","doi":"10.1155/int/3010115","DOIUrl":"https://doi.org/10.1155/int/3010115","url":null,"abstract":"<p>In response to the limitations of traditional Otsu image segmentation—namely high computational overhead and suboptimal segmentation accuracy—this paper proposes an enhanced optimization framework: the Multileader and Fractional-Order Self-learning Sparrow Search Algorithm (MLFS-SSA). By integrating a fractional-order update mechanism, individual agents retain and leverage historical state information during position updates, facilitating escape from local optima and accelerating convergence. Furthermore, a multileader memory strategy is designed to reducing population search pattern redundancy and improving solution diversity. A self-learning mutation mechanism and Lévy flight perturbation are incorporated to reinforce local exploitation and global exploration, respectively. The proposed algorithm is applied to solve the Otsu multithreshold optimization problem, using the interclass variance criterion as the objective function. Extensive experiments on benchmark images demonstrate that MLFS-SSA significantly outperforms classical algorithms in terms of segmentation accuracy and computational efficiency. Ablation studies further confirm the individual contributions of each strategy to the algorithm’s overall performance.</p>","PeriodicalId":14089,"journal":{"name":"International Journal of Intelligent Systems","volume":"2026 1","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-03-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/int/3010115","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147615131","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Artificial Intelligence Framework for Universal Quantification of Surficial Anti-Fingerprint Performance","authors":"Byunghwa Park,&nbsp;Junho Hwang,&nbsp;Sangwook Park","doi":"10.1155/int/3069561","DOIUrl":"https://doi.org/10.1155/int/3069561","url":null,"abstract":"<p>Fingerprint resistance in consumer products is gaining prominence, necessitating reliable anti-fingerprint quantification methods. Traditional approaches include contact angle measurements and image-based methods. However, these often fail under realistic conditions due to transparency and complex textures, limiting accurate fingerprint resistance quantification. Herein, we propose an AI-based framework inspired by human vision that distinguishes fingerprints from complex backgrounds and ranks their visibility from a single photographic image. AW-Net, a novel segmentation tool, achieves 99% cluster accuracy and 93% pixel accuracy on glass surfaces. We introduce a grayscale fingerprint grading metric with five ordinal visibility levels, enabling interpretable, scalable fingerprint resistance quantification closely aligned with human visual perception. The framework establishes its universal applicability for fingerprint resistance quantification across diverse material classes, with 97% cluster accuracy and 90% pixel accuracy on complex and matte wood surfaces, underscoring that its high performance extends beyond relatively simple and glossy glass to more challenging substrates. Overall, by integrating composite visual cues through data augmentation and structured modeling, the framework delivers robust and reliable fingerprint resistance quantification, resolving visibility into five distinct levels across consumer product surfaces.</p>","PeriodicalId":14089,"journal":{"name":"International Journal of Intelligent Systems","volume":"2026 1","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/int/3069561","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147567442","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Meta-Learning Analysis of Deep Neural Network Architectures on Diverse Numeric Datasets via Geometric Complexity Descriptors","authors":"Faruk Bulut,&nbsp;İknur Dönmez","doi":"10.1155/int/8573962","DOIUrl":"https://doi.org/10.1155/int/8573962","url":null,"abstract":"<p>Meta-learning techniques aim to predict the most suitable learning algorithm for a given dataset based on its intrinsic structural characteristics. These techniques provide a robust framework for understanding algorithmic behavior across diverse data distributions and attributes. Although these state-of-the-art models (CNNs and transformers) are widely applied in various machine learning tasks, their use on numerical datasets remains underexplored due to the complexity of their internal structures. This study aims not only to predict the performance of two black-box deep learning models on static datasets but also to conduct a behavioral analysis in order to identify which meta-features most strongly influence their outcomes. It seems unclear which specific attributes of a dataset positively or negatively affect the performance of these deep learning models. To bridge this gap, we constructed a meta-dataset consisting of 296 datasets, each characterized by 20 meta-features describing the dataset’s statistical, geometric, and structural properties. The analysis identifies which intrinsic dataset properties influence model accuracy, without relying on raw data or hyperparameter tuning. Results show that both models perform best on datasets with high feature discriminability, as captured by meta-features such as maximum feature efficiency, collective feature efficiency, and directional separability. In contrast, performance declines with increasing class boundary complexity and nonlinearity, reflected in features like class separability measures and the linear classifier nonlinearity metric. While CNNs are more sensitive to local geometric complexity, transformers respond more strongly to global statistical measures such as mutual information and entropy, highlighting their distinct inductive biases. The proposed meta-model accurately predicts the performance of both architectures on unseen datasets (0.96 correlation coefficient, 0.019 MAE, and 0.025 RMSE for CNNs; 0.92 correlation coefficient, 0.027 MAE, and 0.036 RMSE for transformers), enabling performance estimation without costly training. These findings emphasize the importance of aligning model architecture with dataset geometry and structure. Additionally, the framework supports more interpretable, efficient, and sustainable deep learning model selection in structured data settings.</p>","PeriodicalId":14089,"journal":{"name":"International Journal of Intelligent Systems","volume":"2026 1","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-03-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/int/8573962","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147567134","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}