PeerJ Computer Science最新文献

{"title":"A review of deep learning methods in aquatic animal husbandry.","authors":"Marzuraikah Mohd Stofa, Fatimah Az Zahra Azizan, Mohd Asyraf Zulkifley","doi":"10.7717/peerj-cs.3105","DOIUrl":"10.7717/peerj-cs.3105","url":null,"abstract":"<p><p>Aquatic animal husbandry is crucial for global food security and supports millions of livelihoods around the world. With the growing demand for seafood, this industry has become economically significant for many regions, contributing to local and global economies. However, as the industry grows, it faces various major challenges that are not encountered in small-scale setups. Traditional methods for classifying, detecting, and monitoring aquatic animals are often time-consuming, labor-intensive, and prone to inaccuracies. The labor-intensive nature of these operations has led many aquaculture operators to move towards automation systems. Yet, for an automation system to be effectively deployed, it needs an intelligent decision-making system, which is where deep learning techniques come into play. In this article, an extensive methodological review of machine learning methods, primarily the deep learning methods used in aquatic animal husbandry are concisely summarized. This article focuses on the use of deep learning in three key areas: classification, localization, and segmentation. Generally, classification techniques are vital in distinguishing between different species of aquatic organisms, while localization methods are used to identify the respective animal's position within a video or an image. Segmentation techniques, on the other hand, enable the precise delineation of organism boundaries, which is essential information in accurate monitoring systems. Among these key areas, segmentation techniques, particularly through the U-Net model, have shown the best results, even achieving a high segmentation performance of 94.44%. This article also highlights the potential of deep learning to enhance the precision, productivity, and sustainability of automated operations in aquatic animal husbandry. Looking ahead, deep learning offers huge potential to transform the aquaculture industry in terms of cost and operations. Future research should focus on refining existing models to better address real-world challenges such as sensor input quality and multi-modal data across various environments for better automation in the aquaculture industry.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e3105"},"PeriodicalIF":2.5,"publicationDate":"2025-08-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12453753/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145132412","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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":"Innovative multi objective optimization based automatic fake news detection.","authors":"Cebrail Barut, Suna Yildirim, Bilal Alatas, Gungor Yildirim","doi":"10.7717/peerj-cs.3016","DOIUrl":"10.7717/peerj-cs.3016","url":null,"abstract":"<p><p>With the digital revolution, access to information is expanding day by day and individuals can access information quickly through the internet and social media platforms. However, in most cases, there is no mechanism in place to evaluate the accuracy of news that spreads rapidly on social media. This increases the potential for fake news to mislead both individuals and society. In order to minimize the negative effects of fake news, it has become a critical necessity to detect them quickly and effectively. Metaheuristic methods can provide more effective solutions in fake news detection compared to traditional methods. Especially in small datasets, metaheuristics are known to produce faster and more effective solutions than artificial intelligence and machine learning based methods. In the literature, the majority of fake news detection studies have focused on the optimization of a single criterion. In this study, unlike other studies, a method that enables simultaneous optimization of two criteria (precision and recall) in fake news detection is developed. In the proposed approach, an innovative solution is presented by using the Crowding Distance Level method instead of the Crowding Distance method used in the standard Non-dominated Sorting Genetic Algorithm 2 (NSGA-2) algorithm. The proposed method is tested on four different datasets such as Covid-19, Syrian war daily news and FakeNewsNet (Gossipcop). The results show that the proposed method achieves high success especially on small datasets.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e3016"},"PeriodicalIF":2.5,"publicationDate":"2025-08-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12453838/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145132683","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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":"Detection of offensive content in the Kazakh language using machine learning and deep learning approaches.","authors":"Milana Bolatbek, Moldir Sagynay, Shynar Mussiraliyeva, Zhastay Yeltay","doi":"10.7717/peerj-cs.3027","DOIUrl":"10.7717/peerj-cs.3027","url":null,"abstract":"<p><p>This article addresses the urgent need to detect destructive content, including religious extremism, racism, cyberbullying, and nation oriented extremism messages, on social media platforms in the Kazakh language. Given the agglutinative structure and rich morphology of Kazakh, standard natural language processing (NLP) models require significant adaptation. The study employs a range of machine learning and deep learning techniques, such as logistic regression, support vector machines (SVM), and long short-term memory (LSTM) networks, to classify destructive content. This article demonstrates the effectiveness of combining n-gram and stemming methods with machine learning algorithms, achieving high accuracy in content classification. The findings underscore the importance of developing language-specific NLP tools tailored to Kazakh's linguistic complexities. This research not only contributes to ensuring online safety by detecting destructive content in Kazakh digital spaces, but also provides a framework for applying similar techniques to other lesser-resourced languages.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e3027"},"PeriodicalIF":2.5,"publicationDate":"2025-08-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12453855/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145132568","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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":"DDSUD: dynamically detecting subsequence uncertainty and diversity for active learning in imbalanced Chinese sentiment analysis.","authors":"Shufeng Xiong, Yibo Si, Guipei Zhang, Bingkun Wang, Guang Zheng, Haiping Si","doi":"10.7717/peerj-cs.3091","DOIUrl":"10.7717/peerj-cs.3091","url":null,"abstract":"<p><p>Sentiment structure analysis in Chinese text typically relies on supervised deep-learning methods for sequence labeling. However, obtaining large-scale labeled datasets is both resource-intensive and time-consuming. To address these challenges, this study proposes Dynamically Detecting Subsequence Uncertainty and Diversity (DDSUD), a Bidirectional Encoder Representations from Transformers (BERT)-based active learning framework designed to tackle subsequence uncertainty and enhance the diversity of imbalanced datasets. DDSUD combines subsequence uncertainty detection, diversity-driven sample selection, and dynamic weighting, enabling an adaptive balance between these factors throughout the active learning iterations. Experimental results show that DDSUD achieves performance close to fully supervised training schemes with only 50% of the data labeled, and outperforms other state-of-the-art active learning methods with the same amount of labeled data. Moreover, by dynamically adjusting the trade-off between subsequence uncertainty and diversity, DDSUD demonstrates strong adaptability and generalization capability in low-resource environments, especially in handling imbalanced datasets, significantly improving the recognition of minority class samples.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e3091"},"PeriodicalIF":2.5,"publicationDate":"2025-08-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12453870/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145132476","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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":"Fuzzy evaluation and explainable machine learning for diagnosis of rheumatic and autoimmune diseases.","authors":"Mohammed Fadhil Mahdi, Arezoo Jahani, Dhafar Hamed Abd","doi":"10.7717/peerj-cs.3096","DOIUrl":"10.7717/peerj-cs.3096","url":null,"abstract":"<p><p>In this article, a new combination of an explainable machine learning approach with a fuzzy evaluation framework is proposed to improve the diagnostic performance and interpretation of rheumatic and autoimmune diseases. This work addresses three major challenges: (i) overlapping symptoms and complex clinical presentations, (ii) the lack of interpretability in traditional machine learning models, and (iii) the difficulty of selecting the best diagnosis model. To overcome these challenges, a new dataset was collected from Iraq's hospitals and health centers between 2019 and 2024. The size of dataset is 12,085 patients and includes 14 features in seven classes (rheumatoid arthritis, reactive arthritis, ankylosing spondylitis, Sjogren syndrome, systemic lupus erythematosus, psoriatic arthritis, and normal). The dataset is subjected to extensive preprocessing with attribute imputation (mean and mode), encoding categorical features, and balancing the data to pass it to 12 different machine learning models. Performance is evaluated based on precision, recall, F-score, kappa, Hamming loss, Matthews correlation coefficient, and accuracy to identify the best model. To select the optimal model, we apply fuzzy decision by opinion score method (FDOSM). The FDOSM process involves assessments from three domain experts to ensure a robust and well-rounded evaluation. Furthermore, the explainable artificial intelligence (XAI) technique provides global and local explanations for model predictions. Local interpretable model explanations (LIME) were used as explanations and significantly increased the transparency and reliability of the clinical decision-making process. The results show that the FDOSM yields gradient boosting with a 0.1333 score and a rank of 1, is the best model with an accuracy of 86.89%, precision of 87.35%, and kappa of 84.51%. The best model using XAI to increase confidence and trustworthiness in clinical decision-making and healthcare applications.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e3096"},"PeriodicalIF":2.5,"publicationDate":"2025-08-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12453786/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145132656","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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":"Federated learning for digital twin applications: a privacy-preserving and low-latency approach.","authors":"Jie Li, Dong Wang","doi":"10.7717/peerj-cs.2877","DOIUrl":"10.7717/peerj-cs.2877","url":null,"abstract":"<p><p>The digital twin (DT) concept has recently gained widespread application for mapping the state of physical entities, enabling real-time analysis, prediction, and optimization, thereby enhancing the management and control of physical systems. However, when sensitive information is extracted from physical entities, it faces potential leakage risks, as DT service providers are typically honest yet curious. Federated learning (FL) offers a new distributed learning paradigm that protects privacy by transmitting model updates from edge servers to local devices, allowing training on local datasets. Nevertheless, the training parameters communicated between local mobile devices and edge servers may contain raw data that malicious adversaries could exploit. Furthermore, variations in mapping bias across local devices and the presence of malicious clients can degrade FL training accuracy. To address these security and privacy threats, this paper proposes the FL-FedDT scheme-a privacy-preserving and low-latency FL method that employs an enhanced Paillier homomorphic encryption algorithm to safeguard the privacy of local device parameters without transmitting data to the server. Our approach introduces an improved Paillier encryption method with a new hyperparameter and pre-calculates multiple random intermediate values during the key generation stage, significantly reducing encryption time and thereby expediting model training. Additionally, we implement a trusted FL global aggregation method that incorporates learning quality and interaction records to identify and mitigate malicious updates, dynamically adjusting weights to counteract the threat of malicious clients. To evaluate the efficiency of our proposed scheme, we conducted extensive experiments, with results validating that our approach achieves training accuracy and security on par with baseline methods, while substantially reducing FL iteration time. This enhancement contributes to improved DT mapping and service quality for physical entities. (The code for this study is publicly available on GitHub at: https://github.com/fujianU/federated-learning. The URL address of the MNIST dataset is: https://gitcode.com/Resource-Bundle-Collection/d47b0/overview?utm_source=pan_gitcode&index=top&type=href&;.).</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2877"},"PeriodicalIF":2.5,"publicationDate":"2025-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12453653/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145132515","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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":"LPITutor: an LLM based personalized intelligent tutoring system using RAG and prompt engineering.","authors":"Zhensheng Liu, Prateek Agrawal, Saurabh Singhal, Vishu Madaan, Mohit Kumar, Pawan Kumar Verma","doi":"10.7717/peerj-cs.2991","DOIUrl":"10.7717/peerj-cs.2991","url":null,"abstract":"<p><p>Development of large language models (LLMs) has transformed the landscape of personalized education through intelligent tutoring systems (ITS) which responds to diverse learning requirements. This article proposed a model named LLM based Personalized Intelligent Tutoring System (LPITutor) that is based on LLM for personalized ITS that leverages retrieval-augmented generation (RAG) and advanced prompt engineering techniques to generate customized responses aligned with students' requirements. The aim of LPITutor is to provide customized learning content that adapts to different levels of learners skills and question complexity. The performance of proposed model was evaluated on accuracy, completeness, clarity, difficulty alignment, coherence, and relevance. The finding of LPITutor indicates that it effectively balances the response accuracy and clarity with significant alignment to the difficulty level of student queries. The proposed work also emphasises the broader implications of artificial intelligence (AI)-driven ITS in education and presents future directions for improving the adaptation and optimization of LPITutor.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2991"},"PeriodicalIF":2.5,"publicationDate":"2025-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12453719/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145132704","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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 survey of streaming data anomaly detection in network security.","authors":"Pengju Zhou","doi":"10.7717/peerj-cs.3066","DOIUrl":"10.7717/peerj-cs.3066","url":null,"abstract":"<p><p>Cybersecurity has always been a subject of great concern, and anomaly detection has gained increasing attention due to its ability to detect novel attacks. However, network anomaly detection faces significant challenges when dealing with massive traffic, logs, and other forms of streaming data. This article provides a comprehensive review and a multi-faceted analysis of recent algorithms for anomaly detection in network security. It systematically categorizes and elucidates the various types of datasets, measurement techniques, detection algorithms, and output results of streaming data. Furthermore, the review critically compares network security application scenarios and problem-solving capabilities of streaming data anomaly detection methods. Building on this analysis, the study identifies and delineates promising future research directions. This article endeavors to achieve rapid and efficient detection of streaming data, thereby providing better security for network operations. This research is highly significant in addressing the challenges and difficulties of analyzing anomalies in streaming data. It also serves as a valuable reference for further development in the field of network security. It is anticipated that this comprehensive review will serve as a valuable resource for security researchers in their future investigations within network security.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e3066"},"PeriodicalIF":2.5,"publicationDate":"2025-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12453818/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145132394","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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":"Design of performance evaluation method for higher education reform based on adaptive fuzzy algorithm.","authors":"Dakun Yang, Muhammad Sheraz Arshad Malik","doi":"10.7717/peerj-cs.3090","DOIUrl":"10.7717/peerj-cs.3090","url":null,"abstract":"<p><p>This study presents a performance evaluation framework for university teachers based on the adaptive neural fuzzy inference system (ANFIS), aiming to enhance teaching quality and institutional management through a scientific, objective, and comprehensive assessment mechanism. The proposed method begins by developing a robust evaluation index system that integrates key dimensions of academic activity, including teaching performance, research contributions, and fundamental faculty information. A total of 16 sub-indicators are incorporated into the evaluation framework. To optimize data processing and reduce redundancy, factor analysis is applied, simplifying the indicator set while maintaining the integrity and effectiveness of the evaluation process. The core of the system leverages the strengths of both fuzzy logic and neural networks, combining the capacity of fuzzy systems to handle imprecise and uncertain information with the adaptive learning capabilities of neural networks. This hybrid approach improves the accuracy, interpretability, and adaptability of the evaluation results. By continuously optimizing the model using training data, the system dynamically refines its rule base and parameters, eliminating the reliance on manually defined parameters common in traditional fuzzy systems. The effectiveness of the ANFIS-based evaluation model is validated through empirical experiments. The results demonstrate that the proposed model outperforms conventional methods, such as backpropagation (BP) neural networks and support vector machines (SVMs), in terms of accuracy, precision, and overall performance. This research offers a novel and practical approach for evaluating university teacher performance, enabling more accurate reflection of teaching and research outcomes, and providing valuable decision-making support for academic management.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e3090"},"PeriodicalIF":2.5,"publicationDate":"2025-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12453743/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145132586","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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":"2LE-BO-DeepTrade: an integrated deep learning framework for stock price prediction.","authors":"Zinnet Duygu Akşehir, Erdal Kılıç","doi":"10.7717/peerj-cs.3107","DOIUrl":"10.7717/peerj-cs.3107","url":null,"abstract":"<p><p>This study presents a novel, integrated deep-learning framework named 2LE-BO-DeepTrade for stock closing price prediction. This framework combines 2LE-ICEEMDAN denoising, deep learning models tuned with Bayesian optimization, and a piecewise linear representation (PLR)-based trading strategy. The framework utilizes the model that provides the highest accuracy among optimized long short-term memory (LSTM), long short term memory with batch normalization (LSTM-BN), and gated recurrent unit (GRU) models on data preprocessed with the 2LE-ICEEMDAN denoising method. The model's performance is comprehensively evaluated using both statistical metrics and a PLR-based trading strategy specifically developed for this study. Experimental studies were conducted on AKBNK, MGROS, KCHOL, THYAO, and ULKER stocks, which are traded on Borsa Istanbul and represent different sectors. During the denoising phase, noise in the stock prices was successfully removed, and noiseless intrinsic mode functions (IMFs) were obtained. The optimal model and hyperparameters for each IMF component were determined using Bayesian optimization, significantly improving prediction accuracy. The model within this framework, characterized by its optimized yet simple structure, demonstrated superior predictive performance compared to the more complex ICE2DE-MDL model in the literature. When compared to ICE2DE-MDL, the 2LE-BO-DeepTrade model, across all tested stocks, reduced the average root mean square error (RMSE) value by 94.4%, the average mean absolute error (MAE) value by 93.6%, and the average mean absolute percentage error (MAPE) value by 37.4% while increasing the average R² value by 1.1%. Furthermore, the PLR-based trading strategy, specifically developed for this study, generated \"Buy\" and \"Sell\" signals, exhibiting a remarkably superior financial performance to a passive investment strategy. Across all considered stocks, the PLR-based strategy yielded, on average, 66 times more profit than the passive approach. These findings substantiate that the proposed integrated deep learning-based stock forecasting framework can significantly enhance the accuracy of stock market predictions and the returns of trading strategies.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e3107"},"PeriodicalIF":2.5,"publicationDate":"2025-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12453771/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145132604","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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}