International Journal of Computer Applications in Technology (30 papers in press)

Regular Issues

• Open Access

  Customer churn prediction in live e-commerce based on stacking ensemble learning ( Free Full-text Access )

  by Lijie Shi 
  Abstract: The real-time nature and high noise interactive environment of live streaming e-commerce make it difficult to capture the dynamic behavioral characteristics of customers, which reduces the accuracy of customer churn risk prediction. Therefore, a risk prediction method for customer churn in live streaming e-commerce based on Stacking ensemble learning is proposed. Firstly, detect and eliminate outliers in the data sample, and normalize the data features based on skewness. Secondly, extract customer characteristics of live streaming e-commerce from both time and behavioral dimensions. Finally, using the extracted customer features as input, a live streaming e-commerce customer churn risk prediction model based on Stacking ensemble learning is constructed. The experimental results show that the accuracy of the proposed method is 0.96 and the F1 value is 0.96. Therefore, it indicates that the research method provides accurate customer retention decision support for real-time high noise e-commerce live streaming scenarios.
  Keywords: stacking ensemble learning; live streaming e-commerce; customer churn; risk profile.
  DOI: 10.1504/IJCAT.2026.10077093
   
  

• Open Access

  An accurate prediction method of financial transaction risk based on multidimensional data mining ( Free Full-text Access )

  by Jing Liang 
  Abstract: In order to solve the problems of low F1 value and prediction accuracy, as well as long prediction completion time in traditional financial transaction risk prediction methods, an accurate prediction method of financial transaction risk based on multidimensional data mining is proposed. For multidimensional mining and dimensionality reduction of financial transaction data, the reduced data is input into the WOA-CNN-BiLSTM-Attention model, which initializes BiLSTM with WOA optimized parameters. CNN extracts data features through convolutional and pooling layers, and inputs the extracted feature data into BiLSTM to determine long-term dependencies of the data. Input the features processed by the attention mechanism into the fully connected layer, and output the prediction results through the activation function. The experimental results show that the maximum F1 value of the proposed method is 0.98, the maximum prediction accuracy is 97.23%, and the minimum prediction completion time is 0.78 s.
  Keywords: multidimensional data mining; financial transaction risk; accurate prediction; EWKM; WOA-CNN-BiLSTM-attention model.
  DOI: 10.1504/IJCAT.2026.10077094
   
  

• Open Access

  A multi-dimensional time series algorithm for personalised recommendation in an innovation and entrepreneurship resource library ( Free Full-text Access )

  by Qianqian Shen, Benneng Jin, Xueyin Hong 
  Abstract: To address the limitations of traditional resource library recommendation methods such as low accuracy, prolonged response times and poor user satisfaction this study proposes a personalised recommendation approach for innovation and entrepreneurship resource libraries based on a multi-dimensional time series data algorithm. After obtaining multi-dimensional time series data recommended for the innovation and entrepreneurship resource library and determining the abnormal score of the data, the abnormal data is removed. Extracting spatial and temporal features of multi-dimensional temporal data through graph attention network, fusing the features to determine user and content feature vectors, predicting the correlation score between oral examination users and innovation and entrepreneurship resource library content, and recommending innovation and entrepreneurship resource libraries with high scores to users. The experimental results show that the accuracy of the proposed method varies between 92.38% and 97.89%, with an average response time of 0.76 and an average satisfaction rate of 97.96.
  Keywords: multi-dimensional time series data algorithm; innovation and entrepreneurship; resource library; personalised recommendation; graph attention network.
  DOI: 10.1504/IJCAT.2026.10077095
   
  

• Open Access

  English online interactive learning behaviour recognition based on face detection ( Free Full-text Access )

  by Jing Zhang 
  Abstract: In order to provide real-time feedback and personalized intervention for online English teaching, this paper proposes a facial detection based English online interactive learning behavior recognition method. Firstly, a multi-modal data fusion is used to construct an English online interactive learning behavior standard feature library, with composite constraint conditions and validation criteria set, ultimately forming an interpretable behavior discrimination system. Secondly, the curvature extremum coefficient algorithm was improved by introducing the principal curvature direction derivative and optimizing the preprocessing for outlier removal, which improved the accuracy and robustness of feature point localization. Finally, a functional partitioning strategy based on curvature stability and a dynamic weighted voting mechanism were proposed to achieve fine-grained recognition of learning behavior. The experiment shows that the maximum Jaccard similarity coefficient of this method is 0.92, the maximum Pearson correlation coefficient is 0.94, and the maximum Fowlkes Mallows index can reach 0.93.
  Keywords: English online interactive learning; learning behaviour; standard features; facial detection; behaviour recognition.
  DOI: 10.1504/IJCAT.2026.10077096
   
  

• Open Access

  Intelligent path planning method for UAV inspection of complex distribution network lines under multibranch operation tasks ( Free Full-text Access )

  by Zhifei Zhang 
  Abstract: To address the issues related to poor completion rates, poor planning quality, and long planning time in traditional planning methods for multi branch operation tasks, an intelligent planning method for unmanned aerial vehicle (UAV) inspection path of complex distribution network lines under multi branch operation tasks is proposed Build a multi branch task allocation objective function based on the modeling results of unmanned aerial vehicle inspection environment, and solve the objective function through an improved bidirectional ant colony algorithm to achieve multi branch task allocation The APF-RRT fusion algorithm conducts efficient and orderly search guided by potential field trends to determine the unmanned aerial vehicle inspection path for complex distribution network lines Test results indicate that the suggested method attains a peak completion rate of 99 32% for multi-branch tasks, while ensuring drones avoid danger areas entirely The trajectories remain efficient in length, and the minimum computation time is 1.03 s.
  Keywords: multi-branch operation tasks; complex distribution network lines; unmanned aerial vehicle; inspection path; intelligent planning; APF-RRT fusion algorithm.
  DOI: 10.1504/IJCAT.2026.10077097
   
  

• Open Access

  An automatic recommendation method for online teaching resources based on SA-BPNN multimodal fusion ( Free Full-text Access )

  by Jinmei Lv 
  Abstract: To address the challenges of cross-modal feature fusion and semantic association in online education, this study develops an intelligent resource recommendation method based on SA-BPNN multimodal fusion. The proposed approach first employs a self-attention mechanism to dynamically weight features from video, exercise, and behavioral modalities, establishing an "input-process-output" correlation model. A hybrid fusion layer is then applied to handle potential missing modalities, with BPNN optimizing the integrated feature representation. Furthermore, a structured teaching knowledge graph is constructed, and resource similarity is computed by combining user-resource rating matrices with semantic representations from the graph, ultimately generating a Top-N recommendation list. Test results demonstrate that the method maintains stable performance, with coverage remaining between 97.68% and 99.45%, and P/N measure consistently exceeding 0.8
  Keywords: SA-BPNN multimodal fusion; online teaching resources; automatic recommendation; knowledge graph.
  DOI: 10.1504/IJCAT.2026.10077220
   
  

• Open Access

  Chinese named entity recognition with a residual convolutional neural network ( Free Full-text Access )

  by Pingge Huang, Xinhua Wang 
  Abstract: In order to improve the accuracy and F1 score of Chinese named entity recognition, this paper proposes a Chinese named entity recognition method based on residual convolutional neural network and dynamic adaptive activation mechanism. Firstly, residual connections and Dynamic ReLU activation strategies are introduced to address the challenges of gradient vanishing and network degradation faced by deep networks. Subsequently, to enhance the ability of LSTM in short message fitting, residual connection module, Dynamic ReLU module, and attention regulation unit were added. Finally, by integrating nonlinear connections, residual units with adaptive activation characteristics and Dynamic ReLU, as well as attention mechanisms that dynamically adjust network depth, the method effectively alleviates the performance degradation problem in deep networks, thereby improving the efficiency of Chinese named entity recognition. Experimental results show that our method consistently achieves over 96% accuracy and an F1 score above 0.94.
  Keywords: residual convolutional neural network; Chinese named entity recognition; dynamic ReLU activation mechanism; feature vector.
  DOI: 10.1504/IJCAT.2026.10077221
   
  

• Open Access

  An infrared thermography-based study on temperature anomaly identification for electrical cabinets in offshore wind farm substations ( Free Full-text Access )

  by Xiaoyan Liu, Faqing Wei, Cimin Su, Lin Zhang 
  Abstract: The offshore wind farm booster station has been in a high salt spray marine environment for a long time, which makes it difficult to detect abnormal temperature in the electrical cabinet. Therefore, a method for identifying temperature anomalies in electrical cabinets of offshore wind farm boosting stations based on infrared temperature measurement is proposed. Firstly, infrared temperature measurement technology is used to accurately monitor the surface temperature of the electrical cabinet. Secondly, by using deep learning architecture to achieve multidimensional feature analysis of infrared thermal images, the recognition accuracy of temperature anomaly areas can be improved. Finally, based on three-dimensional spatial operations and dynamic segmentation algorithms, the classification and identification of nonlinear features of the temperature field in the electrical cabinet are completed. The results show that the temperature identification error of the method proposed in this paper is relatively small, and the accuracy of temperature anomaly detection reaches 92.9%.
  Keywords: infrared temperature measurement; offshore wind farms; boosting station; electrical cabinet; temperature anomaly identification.
  DOI: 10.1504/IJCAT.2026.10077222
   
  

• Open Access

  Prediction of SF6/N2 decomposition gas concentration based on infrared spectroscopy data ( Free Full-text Access )

  by Guodong Li, Wang Xuan, Yanpeng Li, Qingsong Chen, Xiaohong Zhao 
  Abstract: In order to overcome the problems of low signal peak signal-to-noise ratio, low accuracy, and long time consumption in traditional SF6/N2 decomposition gas concentration prediction methods, this study proposed a prediction method of SF6/N2 decomposition gas concentration based on infrared spectroscopy data. Collecting gas absorption spectrum data through infrared spectroscopy technology, optimizing variational mode decomposition parameters using classification particle swarm optimization algorithm, and achieving noise suppression and feature enhancement of infrared spectrum data signals. Combining the artificial fish swarm algorithm to optimize the initial weights and thresholds of the BP neural network, the processed infrared spectral data is used as input to build a pre built AFSA-BP prediction model, and relevant prediction results are obtained. The experimental results show that the PSNR value of the infrared spectral signal of the proposed method always remains above 47 dB, with the highest accuracy reaching 98.9% and the lowest time consumption being 0.76 s.
  Keywords: infrared spectroscopy data; SF6/N2; decompose gas; concentration prediction; artificial fish swarm algorithm; BP neural network.
  DOI: 10.1504/IJCAT.2026.10077223
   
  

• Open Access

  Action-guided masking strategy for recognising foul moves in track and field athletes ( Free Full-text Access )

  by Lingqiang Xuan, Wei Wang, Di Zhang 
  Abstract: This paper proposes a novel two-stage framework for automated foul recognition in track and field. The first stage introduces a self-learning feature pyramid network (FPANet), which incorporates a multidimensional self-learning attention module (MSLM) and a channel separation extraction module (CSEM) to enhance human keypoint detection. This design addresses the underutilization of multi-resolution features and inter-keypoint feature coupling in complex motion scenarios, improving detection accuracy and robustness. The second stage employs an innovative Action Guided Masking Strategy to dynamically select informative keyframes while suppressing redundant and background frames. This forces the model to infer complete actions from partial observations and sharpens its ability to discern subtle foul movements. An auxiliary classifier further refines the encoded features for final classification. Experimental results confirm the framework's superior performance in both keypoint detection and foul recognition, showcasing its strong potential for intelligent sports analytics.
  Keywords: keypoint detection; foul move recognition; action-guided masking; FPANet; feature pyramid attention network.
  DOI: 10.1504/IJCAT.2026.10077357
   
  

• Open Access

  Study on leveraging fuzzy neural networks for error labelling in English machine translation ( Free Full-text Access )

  by Keli Hu 
  Abstract: Accurate error labelling in English machine translation (MT) is crucial for enhancing translation quality and advancing natural language processing technologies. To address this need, this paper proposes a novel error labelling method based on a fuzzy neural network (FNN). The process begins by collecting MT data using the Scrapy framework and extracting a set of surface and grammatical features. These features are then normalised via Min-Max scaling and concatenated into fixed-order continuous vectors. The core of our method involves optimising the FNNs fuzzy rule connection weights and bias vectors with a Bat Algorithm, enabling the network to perform the final error labelling. Experimental results demonstrate the superior performance of our approach, which achieves a minimal average error omission rate of 0.19%, a high average Kappa coefficient of 0.95, and a fast average labelling time of only 1.5 seconds.
  Keywords: English machine translation; feature; fuzzy neural network; bat algorithm.
  DOI: 10.1504/IJCAT.2026.10077358
   
  

• Open Access

  Supply chain demand forecasting and low-altitude economic risks based on machine learning and artificial intelligence ( Free Full-text Access )

  by Jilu Liu, Yubin Ying, Shaohui Shu, Zechen Zhang 
  Abstract: To address the challenges of coupling high-frequency supply chain forecasting with low-altitude economic risk, a unified framework integrating machine learning and artificial intelligence is proposed. A Long Short-Term Memory (LSTM)-Transformer hybrid model is designed to capture multi-granularity temporal dependencies and mitigate the bullwhip effect. A Graph Attention Network (GAT)-driven risk coupling network is constructed, integrating AirSim simulations with real warehouse data for dynamic quantification. Results show that the forecast RMSE is reduced by 35.3% compared to the Autoregressive Integrated Moving Average (ARIMA) model, and the multi-granularity MAPE variance is only 0.123. Through a joint decision-making mechanism, this method provides an intelligent decision-making paradigm for multi-source heterogeneous scenarios.
  Keywords: supply chain demand forecasting; low-altitude economy; risk analysis; machine learning; artificial intelligence.
  DOI: 10.1504/IJCAT.2026.10077901
   
  

• Open Access

  Ecological management algorithm of fresh supply chain information under artificial intelligence and blockchain architecture ( Free Full-text Access )

  by Chanjuan Li 
  Abstract: This paper establishes a new supply chain management system based on artificial intelligence and blockchain technology (BT), and proposes a new supply chain solution on this basis. This paper defines the traditional centralised management mode as Supply Chain 1 and the blockchain distributed architecture as Supply Chain 2. Sample 1 consists of 500 sets of structured transaction data, while Sample 2 consists of 500 sets of unstructured sensor log data. The results showed that in sample 1, the information transmission times of supply chain 1 and supply chain 2 were 15.4 s and 3.3 s. In sample 2, when the sample size was 500, the information transmission times of SC 1 and SC 2 were 20.7 s and 4.2 s. It can be seen that SC 2 has a higher information transmission efficiency. This evaluation algorithm can be used in practice to diagnose supply chain information transmission bottlenecks and optimise resource allocation.
  Keywords: artificial intelligence; blockchain architecture; fresh supply chain; information ecological management.
  DOI: 10.1504/IJCAT.2026.10077983
   
  

• Open Access

  Simulation model of marine biological population based on multi-agent comfort drive ( Free Full-text Access )

  by Xi Deng, Wei Liang, Yupeng Zhu 
  Abstract: Current simulation models that focus on marine biological communities have not yet taken into account the effects of comfort on the behaviour of single organisms. This shortcoming makes it difficult for researchers to develop accurate simulations of how groups behave when faced with adaptive behaviour in complicated environments. A Multi-Agent System based on Reinforcement Learning (RL) is proposed in this paper, which uses Environmental Perception Mechanisms and Comfort Functions to optimise decisions made by individual agents in complex marine environments via Q-Learning and Deep Q-Networks (DQN). The proposed approach employs a collaborative strategy that combines local and global comfort to simulate collective actions. Experimental results indicate that a comfort-driven decision-making model would enable a collective of individuals to converge within 240 seconds, with 85% of the individuals congregating in areas they perceived as having the highest comfort levels. This work opens up new avenues for simulating the behaviours of marine biological communities and demonstrates the potential of a comfort-driven approach to studying the adaptive behaviour of groups.
  Keywords: marine species group simulation; comfort-driven mechanism; multi-agent system; reinforcement learning; deep Q-network.
  DOI: 10.1504/IJCAT.2026.10078256
   
  

• Open Access

  CISAA: a collective intelligence-based optimisation framework for low-altitude airspace resource allocation and engineering management ( Free Full-text Access )

  by Feng Li 
  Abstract: The rapid growth of the low-altitude economy and complex airspace operations pose challenges for efficient and adaptive resource allocation. Traditional methods struggle with multi-dimensional coupling among resources, missions, and management under uncertainty. This paper proposes a Collective Intelligence-based Spatio-Allocation and Administration (CISAA) framework that integrates swarm intelligence with hierarchical dynamic optimization to enable collaborative decision-making among UAVs, infrastructure units, and management systems. Through multi-layer coordination, CISAA optimizes task scheduling, airspace utilization, and engineering progress. Experiments in simulated low-altitude scenarios show that CISAA achieves higher resource efficiency, task completion, and system stability compared with heuristic and centralized approaches, offering a novel path toward intelligent and sustainable low-altitude airspace management.
  Keywords: low-altitude airspace management; collective intelligence optimisation; multi-agent coordination; engineering resource allocation; adaptive scheduling.
  DOI: 10.1504/IJCAT.2026.10078501
   
  

• Open Access

  Study on global control of solar power generation by integrating maximum power point tracking ( Free Full-text Access )

  by Yiming Peng, Hanzhi Zhang, Chunkai Zhang, Huapeng Shan, Qian Lv 
  Abstract: A global control strategy integrating maximum power point tracking (MPPT) is proposed to optimize solar power generation under complex lighting conditions. Based on the photovoltaic cell equivalent circuit model, the nonlinear relationship between output power, temperature, and light intensity is derived, and the characteristics of series-parallel arrays are analyzed. For uniform conditions, efficient MPPT is achieved by combining the optimal gradient method with a variable step-size strategy. Under non-uniform illumination, a global control method based on the maximum-minimum ant colony algorithm is introduced, transforming multi-peak MPPT into a global optimization problem. Experimental results demonstrate that the proposed strategy provides effective theoretical support and a technical framework for the intelligent control of solar energy systems
  Keywords: Maximum power point tracking; Solar power generation capacity; Global control; Ant algorithm.
  DOI: 10.1504/IJCAT.2026.10079099
   
  

• Open Access

  Semantic-guided neural network for complex dance action recognition ( Free Full-text Access )

  by Lingjie Kong 
  Abstract: The recognition of complex dance movements poses challenges to existing methods based on local features and shallow temporal modeling due to the nonlinearity of movements, multi-scale rhythm changes, and complex background interference. Therefore, a dance action recognition framework based on semantic guided neural network is proposed. Firstly, precise extraction and tracking of dance foreground can be achieved through contour calibration and periodic motion modeling. Secondly, design a semantic guided encoder decoder network that preserves action details through skip connections and innovatively introduces a global spatiotemporal Transformer module to capture dynamic associations between actions. Finally, by combining the multi-scale spatiotemporal relationship learning module and integrating cross scale semantic information, the model's ability to represent complex dance movements is significantly enhanced. The experimental results show that the proposed method maintains a stable accuracy of over 90% in temporal classification, with a single frame processing delay controlled within 30-33ms.
  Keywords: semantic guided neural network; complex dance movements; action recognition; contour calibration.
  DOI: 10.1504/IJCAT.2026.10079100
   
  

• Open Access

  Application of information optics experimental methods in optical sensors for wearable devices ( Free Full-text Access )

  by Yifang Li, Yingji He 
  Abstract: Traditional optical sensors struggle with high-precision physiological monitoring in dynamic environments and lack sufficient processing power for real-time wearable devices. This paper presents an information optics method using a microlens array-based optical sensor with advanced optical modulation to enhance resolution and dynamic response. A combined FM and AM signal transmission improves anti-interference and data speed. A filtering-based data processing method and real-time feedback mechanism ensure signal accuracy and timeliness. Experimental results show that under low, medium, and high noise, the transmission rate increases by 350 bps, 404 bps, and 281 bps, respectively, while bit error rates drop to 0.01, 0.03, and 0.07. In dynamic measurements of heart rate, body temperature, blood oxygen saturation, and blood pressure, accuracy exceeds 92%, and average response time is 0.20 seconds. This demonstrates the effectiveness of the information optics method in wearable optical sensors.
  Keywords: information optics experiment; wearable devices; optical sensor; microlens array; filtering algorithm.
  DOI: 10.1504/IJCAT.2026.10079101
   
  
• FPGA implementation and Multisim simulation of a new four-dimensional two-scroll hyperchaotic system with coexisting attractors  
  by Sundarapandian Vaidyanathan, Esteban Tlelo-Cuautle, Khaled Benkouider, Aceng Sambas, Ciro Fabian Bermudez-Marquez, Samy Abdelwahab Safaan 
  Abstract: Field-programmable gate array (FPGA) design of a new four-dimensional two-scroll hyperchaotic system is investigated in this work. A detailed system modelling of the new system with a hyperchaotic attractor begins this work with phase plots, which is followed by a bifurcation study of the new system. Special dynamic properties such as multistability and symmetry are also investigated for the new system. Using Multisim software, a circuit model is designed and simulated for the new hyperchaotic system. FPGA design and Multisim simulation of the new system enable practical applications in science and engineering. The implementation of the FPGA design in this work is carried out by applying two numerical schemes, viz. Forward Euler and Trapezoidal methods. Experimental attractors observed in the oscilloscope show good match with the Matlab signal plots.The FPGA hardware resources are detailed for both numerical methods.
  Keywords: hyperchaos; bifurcation; symmetry; phase plots; hyperchaotic system;rnparameters; stability; multistability; circuit model; FPGA implementation.
  
  
• Improving hybrid-layer convolutional neural network system for lung cancer nodule classification using enhanced weight optimisation algorithm  
  by Vikul Pawar, P. Premchand 
  Abstract: In recent times, lung cancer is evolving as a highly life-threatening disease for human beings. According to the WHO, lung cancer disease is the second largest cause of deaths as compared to all other types of cancer. The prevailing available technology is striving to get more exposure in the field of medical science using Computer Assisted Diagnosis (CAD), where image processing is playing a crucial role for detecting the cancerous nodules in computer tomographic images. Augmenting the machine learning techniques with image processing algorithms is becoming a more comprehensive examination of cancer disease in proposed CAD systems. This paper is describes a heuristic approach for lung cancer nodule detection, and the proposed model predominantly consists of the following tasks, which are image enhancement, segmenting ROI (Region of Interest), features extraction, and nodule classification. In pre-processing, primarily the Adaptive Median Filter (AMF) filtering method is applied to eliminate the speckle noise from input CT images of Lung Image Database Consortium (LIDC) and Image Database Resource Initiative (IDRI): in the LIDC-IDRI dataset, the quality of input image is improved by applying Histogram Equalization (HE) technique with Contrast-Limited Adaptive (CLA) approach. Secondly, in the successive stage the Improved Level-Set (ILS) algorithm is used to segment the ROI. Furthermore, the third step of the projected work is applied to extract the definite learnable texture features and statistical features from the segmented ROI. The extracted features in the subsequent stage of classification are applied to Hybrid-Layer Convolutional Neural Network (HL-CNN) architecture to classify the lung cancer nodule as either benign or malignant. Principally this research is carried out by contributing to each stage of it, where the novel concept of the improved Hybrid-Layer Convolutional Neural Network (HL-CNN) is employed by optimising and selecting the optimal weight using the Enhanced Cat Swarm Optimisation (ECSO) algorithm. The experimental result of the proposed HL-CNN using the weight optimisation algorithm ECSO is achieved an accuracy of 93%, which is comparatively efficient with respect to existing models such as DBN, SVM, CNN, WOA, MFO, and CSO. Moreover, the proposed model conclusively gives a decision on the detected nodule as either benign or malignant.
  Keywords: Computer Assisted Diagnosis (CAD); Computer Vision; Cancer Diagnosis; Image Classification; Image Enhancement; Image Segmentation; Feature Extraction.
  
  
• Prediction model for total amount of coke oven gas generation based on FCM-RBF  
  by Lili Feng, Jun Peng, Zhaojun Huang 
  Abstract: The rational use of Coke Oven Gas (COG) is of great significance to improve the economic efficiency of enterprises. In this paper, a COG generation prediction model based on fuzzy C-mean clustering (FCM) and radial basis function (RBF) neural network is proposed to address the problems such as the difficulty of accurate modelling of COG generation process and the difficulty of real-time flow prediction. Firstly, the coke oven production process is analysed and correlation analysis is used to select the influencing factors. Secondly, the FCM is used to classify the working conditions of the coke oven, and the appropriate number of working conditions is selected through experiments. Finally, the prediction models under different working conditions are established separately by using RBF. The experiments were carried out using actual industrial production data, and the experimental results showed that the model could provide guidance reference for the dispatchers.
  Keywords: coking oven process; fuzzy C-means clustering; prediction model; radial basis function neural network.
  
  
• Unmanned aerial vehicle non-stop inspection and obstacle avoidance route planning based on reinforcement learning cuckoo search algorithm  
  by Xiaoyong Han, Weiling Lv 
  Abstract: In order to improve the success rate and coverage of obstacle avoidance routes for unmanned aerial vehicle (UAV) during non-stop inspections, a UAV non-stop inspection obstacle avoidance route planning method based on reinforcement learning cuckoo search algorithm is proposed. Firstly, establish a mathematical model that comprehensively considers environmental information, inspection task requirements, and drone performance constraints. Secondly, build a drone non-stop obstacle avoidance inspection route planning model that integrates capability limitations and operational requirements. Finally, the cuckoo search algorithm is used to solve the obstacle avoidance route planning for drone inspection, and reinforcement learning is used to optimize the search mode. Improved algorithm combined with Q-learning, enhancing global search efficiency through three strategic actions to improve solving efficiency and stability. Test results show that the proposed method consistently maintains a success rate of over 97% in obstacle avoidance during drone inspections, and the route coverage rate remains consistently above 95%.
  Keywords: reinforcement learning; cuckoo search algorithm; unmanned aerial vehicle; non-stop inspection; obstacle avoidance route planning.
  DOI: 10.1504/IJCAT.2025.10077513
   
  
• Personalised recommendation methods for information resources on English online learning platforms  
  by Jing Liang 
  Abstract: In this paper, a new personalised recommendation methods for information resources on English online learning platforms is proposed. Design ER storage scale and keyword storage scale, and complete the construction of the English online education platform database. Based on data from English online education platforms, overlapping community partitioning is achieved through Bayesian non negative matrix decomposition, and user interest labels are automatically annotated to achieve user group modeling. Combining user interest preference features and cognitive level features, personalised English resource recommendations are completed. Through experiments, the results show that the proposed method has a recommendation accuracy of up to 95.4%, a recall rate of up to 94.1%, and a recommendation time consistently below 0.91s, demonstrating good application effectiveness.
  Keywords: English; online learning platform; information resources; personalised recommendation.
  DOI: 10.1504/IJCAT.2025.10077514
   
  
• Slope stability assessment based on a combined weight-cloud model  
  by Decheng Zhang, Jielu He, Yunjun Yang 
  Abstract: To evaluate the stability of slopes accurately and reduce the risk of landslide accidents during slope treatment, this study focuses on four aspects: slope morphology, rock mass characteristics, geological structure, and triggering factors. Based on these factors and the intrinsic relationships among indicators, the study refines them into 12 secondary indicators to establish a risk assessment system for slope instability. This research aims to address limitations in previous studies, where the combined influence of subjective judgment and objective data was not fully integrated, and uncertainties in slope evaluation indicators were not effectively managed. To improve the accuracy of weighting, a combination of subjective weighting and objective weighting methods is adopted. The cloud model is used to solve the uncertainty and imprecision of the impact evaluation index, and the slope stability grade is determined according to the principle of maximum certainty. Finally, the method is applied to an engineering example, and the slope stability grade obtained is consistent with the actual slope stability grade, demonstrating that this method can accurately evaluate slope stability.
  Keywords: slope instability; stability evaluation; combined weighting; cloud model.
  DOI: 10.1504/IJCAT.2025.10078410
   
  
• Personalized recommendation of multi-objective tourism routes based on improved shuffled frog leaping algorithm  
  by Weina Pei, Fengjuan Tian 
  Abstract: To address the limitations of low accuracy, long computation time, and low user satisfaction associated with traditional multi-objective tourism route personalised recommendation methods, this study proposes a novel personalised recommendation method based on an improved shuffled frog leaping algorithm. Tourist attraction data were compiled using the Scrapy framework. The compiled data were then clustered via spectral clustering and combined with relevant data to achieve regional division of attractions using an ARMA model. A personalised recommendation model for multi-objective tourism routes was constructed based on the divided attraction regions, incorporating multiple constraint conditions. An improved shuffled frog leaping algorithm was employed to solve the recommendation model, yielding the optimal personalised recommendation scheme. Experimental results demonstrate that the proposed method achieves a maximum recommendation accuracy of 97.4%, a minimum recommendation time of 0.98 s, and a high user satisfaction score of 9.67.
  Keywords: improved shuffled frog leaping algorithm; multi-objective; tourism routes; personalised recommendation; spectral clustering algorithm; regional division of tourist attractions.
  DOI: 10.1504/IJCAT.2025.10078500
   
  
• Signal distortion schemes for peak-to-average power ratio reduction over multi-carrier based orthogonal frequency division multiplexing system: a survey  
  by Parag Jain, Shruti Dixit 
  Abstract: TThe rapid development of multimedia applications necessitates the ability to transmit large volume of data with certain mobile capabilities of multi-carrier based orthogonal frequency division multiplexing (OFDM) communication systems. Increased data rates and high mobility in the OFDM system raise the number of subcarriers, which in turn, causes a major problem of high peak-to-average power ratio (PAPR). The aim of this survey is to give the readers and practitioners in the fields wider understanding of schemes for signal distortion for minimising PAPR problem in OFDM systems and make a taxonomy of such systems. A taxonomy of signal distortion scheme is classified into four techniques namely companding, peak windowing, clipping & filtering and peak cancellation. The metric of this survey for evaluating the same techniques is based on complexity analysis and performance associated factors such as transmitted power, error rates, bandwidth expansion etc. Moreover, this survey also focuses on the advantages, disadvantages and use case suitability for the above-mentioned techniques which helps readers to make a proper judgment about the selection of techniques corresponding to the given scenario.
  Keywords: clipping & filtering; peak windowing; orthogonal frequency division multiplexing; companding; peak cancellation; peak-to-average power ratio.
  DOI: 10.1504/IJCAT.2025.10078755
   
  
• Monitoring emerging trends by content correlation in social media channels using time-stamping technique of blockchain technology  
  by Fazal Tariq, Muhammad Tufail, Taj Rehman 
  Abstract: The rapid evolution of social media demands novel techniques for real-time trend detection on decentralised platforms. Traditional methods such as ARIMA, machine learning, NLP and network analysis face limitations due to data silos, lack of transparency and scalability constraints. This work proposes a TST-based blockchain scheme for content correlation and trend tracking, utilising Ethereum smart contracts to generate tamper-evident timestamps for cross-platform synchronisation. Combined with BERT-based semantic embeddings and temporal metadata, our solution enables transparent, auditable, low-latency trend detection. Experiments on 10,000 posts from Twitter, Reddit and Facebook show that TST achieves 81.2% accuracy and an F1-score of 0.83, outperforming ARIMA by 33.9%, traditional ML by 20.828.7%, NLP by 35.2% and network methods by 34.2%. With a latency of just 2.0 seconds, TST is 912
  Keywords: SMA; social media analysis; trend analysis; content correlation; BCT; blockchain technology; decentralised data; machine learning; TST; time-stamping technique.
  DOI: 10.1504/IJCAT.2025.10079036
   
  
• Effective product recommendation system through topic modelling and K-means approach  
  by Manu G. Thomas, S. Senthil 
  Abstract: Product recommendation systems are widely used to deliver personalised suggestions based on user preferences. However, traditional approaches often suffer from high computational complexity, overfitting, and limited accuracy when handling large-scale textual data. This study proposes an effective product recommendation system using topic modelling and clustering techniques. Latent Dirichlet Allocation (LDA) combined with term frequency inverse document frequency (TF-IDF), count vectorisation, and doc2vec is employed to extract meaningful topics from Twitter data. A modified k-means clustering approach, incorporating arithmetic mean-based centroid initialisation and cosine similarity, improves clustering accuracy and convergence. The model utilises Twitter event data and Flipkart product review datasets to generate relevant recommendations. Experimental results demonstrate that the proposed TF-IDF-based approach outperforms existing methods in terms of similarity score and accuracy, providing improved recommendation quality and effective handling of high-dimensional textual data.
  Keywords: product recommendation; recommendation systems; topic modelling; latent Dirichlet allocation; LDA; term frequency inverse document frequency; TF-IDF; count vectorisation; doc2vec; k-means clustering; cosine similarity; Twitter data; Flipkart dataset; machine learning.
  DOI: 10.1504/IJCAT.2025.10079102
   
  
• Anomaly detection for English MOOC teaching platform data based on improved random forest  
  by Hongmei Niu 
  Abstract: In order to solve the problem of low AUC value and Kappa coefficient in traditional detection methods, an anomaly detection method for English MOOC teaching platform data based on improved random forest is proposed. Firstly, the raw data is cleaned and processed using approximate symbol aggregation algorithm to eliminate outliers such as bias, missing, and redundancy; Then, the cleaned data features are extracted through the autoencoder component in the stack model architecture; Finally, the confusion factor and decision tree maximum depth constraint factor are introduced to improve the random forest algorithm, and the improved algorithm is used to detect data anomalies. The experimental results show that the maximum AUC value of this method can reach 0.981, and the upper and lower limits of Kappa coefficient are 0.94 and 0.85, respectively, indicating that this method can effectively distinguish between normal data and abnormal data.
  Keywords: English MOOC teaching platform; data; anomaly detection; random forest; autoencode; data features; feature confusion factor; maximum depth constraint factor.
  DOI: 10.1504/IJCAT.2025.10079103
   
  
• HLR algorithm: a novel approach for mathematical expressions recognition from ruled handwritten data samples  
  by Sakshi Arora, Chetan Sharma 
  Abstract: The recognition of handwritten mathematical expressions has been a challenging problem, particularly emphasized during the shift to online education due to COVID-19. Researchers conducted an experiment using a dataset of handwritten expressions collected from 991 students on regular ruled sheets, aiming to develop recognizers and evaluators for assessing students' exam papers in scientific and mathematical subjects. The study introduces a novel horizontal line removal (HLR) algorithm for effective document preprocessing. Following this, a neural network-based recognizer was trained on the dataset, with various neural network architectures tested to achieve optimal performance. The best model reached a recognition accuracy of 87.89%. This research is significant as it is the first to focus on mathematical expressions written on ruled sheets, offering a fresh perspective that could enhance the future of educational assessment and technology integration in learning environments.
  Keywords: mathematical expressions; pre-processing; segmentation; neural network; handwritten mathematical expressions; recognition.
  DOI: 10.1504/IJCAT.2025.10079104