Digital Library

Vol. 21, No. 3, Jun. 2025  Cover

1. Water Quality Prediction Model Based on Tucker Decomposition and GRU-Attention Mechanism
   Xuegang Luo, Junrui Lv, Hongrui Yu, Juan Wang
   Vol. 21, No. 3, pp. 227-239, Jun. 2025  
   https://doi.org/10.3745/JIPS.04.0346
   Details Full Text PubReader
   Keywords: Encoder-Decoder Network, Tensor decomposition, Water Quality Prediction
   Show / Hide Abstract

   The issue of water pollution critically affects all living beings. The implementation of a smart water quality monitoring system, based on the Internet of Things, enables advancements in efficiency, security, and cost-effectiveness while providing real-time capabilities. Current water quality prediction models often fail to fully utilize data characteristics shared by water quality indicators, resulting in poor predictive accuracy. This study introduces a novel water quality prediction model named TGMHSA, which utilizes tensor decomposition combined with a gated neural network and a multi-head self-attention mechanism. The aim is to tackle the difficulty of forecasting water quality indicators using time series data while minimizing the risk of plagiarism. The proposed model utilizes standard delay embedding transformation (SDET) to convert the time series data into tensor data, extracting data characteristics by Tucker tensor decomposition, and then combines a multi-head self-attention mechanism to discover potential relationships among data characteristics of multiple water quality indicators. Finally, the utilization of the GRU model enables accurate prediction of multi-index water quality. In order to compare its performance, we consider four indices: root mean square error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE), and coefficient of determination represented as R2. The outcomes demonstrate that this model outperforms traditional methods for predicting water quality in terms of accuracy and resilience, thereby establishing a scientific foundation for effective water quality prediction and environmental monitoring management.


2. A Novel Approach to Manage Sewage Treatment Plants Based on Process Mining Technology
   Jingbo Zhao, Bin Shao, Hao Jiang, Chao Liu, Sheng Miao
   Vol. 21, No. 3, pp. 240-254, Jun. 2025  
   https://doi.org/10.3745/JIPS.04.0347
   Details Full Text PubReader
   Keywords: Conformance Checking, Data Mining, Management, Petri Net, Process Mining
   Show / Hide Abstract

   Due to the complexity of the operation process in sewage treatment plants, there are numerous potential risks involved in the process. An appropriate business process model is necessary for effective staff management and risk detection. However, conventional modeling methods are inherently subjective in the field of sewage treatment. Designers not only have to grasp the workflow language but also need to be familiar with the whole business process. Compared to conventional data mining, process mining specializes in end-to-end processes. Consequently, process mining is better adapted to solving process problems. In this paper, a novel approach is proposed to analyze the operational risk of sewage treatment plants by using process mining technology. The ideal Petri net model and event logs are utilized for conformance checking. The results of the experiment indicate that the proposed approach can discover operational processes from event logs in the field of sewage treatment. The method can assist managers detect staff deviating from standard operating procedures. The results of the implemented process mining technology present the sewage treatment plant managers with a real analysis and understanding to make the staff’s operation easier. The methodology can be extended to similar scenarios.


3. A Improved A* Algorithm for the Path Planning Problem of Urban Taxi-Carpooling
   Qiang Xiao, Guoqing Song, Ziyi Wang
   Vol. 21, No. 3, pp. 255-270, Jun. 2025  
   https://doi.org/10.3745/JIPS.04.0348
   Details Full Text PubReader
   Keywords: Improved A* Algorithm, Path Planning, Taxi Carpool, Traffic Road Network
   Show / Hide Abstract

   In order to address the issue of taxi carpooling path planning on urban roads, this study suggests an improved A* algorithm and a model based on node weight. The path planning model enables us to implement carpool path planning after carpool passengers, taxi passengers, and taxi drivers have gathered. It uses a vector city traffic road network, city road vector map topology, dynamic road weight functions, node weight tables of the road, and an improved A* algorithm. Our evaluation of the model involves comparing its path planning computation time and total travel time with the traditional A* algorithm using Nanjing taxi trajectory data. The comparison shows that the proposed algorithm significantly outperforms the traditional A* algorithm. Results show that the taxi path planning model proposed in this paper can provide a reference for carpool passengers, taxi passengers, and taxi drivers in choosing a carpool.


4. Toward Automatic Parallelization: Detecting Parallelizable Loops Using Large Language Modeling
   Soratouch Pornmaneerattanatri, Keichi Takahashi, Yutaro Kashiwa, Kohei Ichikawa, Hajimu Iida
   Vol. 21, No. 3, pp. 271-283, Jun. 2025  
   https://doi.org/10.3745/JIPS.04.0349
   Details Full Text PubReader
   Keywords: Automatic Parallelization, Deep Learning, Large Language Model, OpenMP, Parallel computing
   Show / Hide Abstract

   To fully harness the capabilities of multi-core processors, parallel programming is indispensable, demanding a comprehensive understanding of both software and hardware. Although various tools have been developed to automate parallel programming by leveraging static analysis approaches, manually parallelized code continues to consistently outperform those that are automatically parallelized. Meanwhile, the emergence of transformer-based large language models has facilitated significant breakthroughs in understanding and generating programming languages. This study presents a model tailored to detecting parallelizable for-loops by fine-tuning the transformer-based model, CodeT5. The fine-tuned model assists programmers in identifying independent for-loops that have the potential for parallelization with libraries like OpenMP, leading to performance enhancements in software applications. The model was trained on 500,000 for-loops sourced from public GitHub repositories and demonstrated an F1-score of 0.860 in detecting parallelizable for-loops within a public GitHub dataset, along with an F1-score of 0.764 on the NAS Parallel Benchmark suite.


5. The Application of DCRN Algorithm Based on Attention Mechanism for Classification in Media Images
   Peipei Dai
   Vol. 21, No. 3, pp. 284-295, Jun. 2025  
   https://doi.org/10.3745/JIPS.02.0224
   Details Full Text PubReader
   Keywords: Attention Mechanism, Convolutional Neural Network, DCRN, Media Images, Multi-label classification
   Show / Hide Abstract

   With the development of the information society, the significant increase in images in online environments poses challenges to media image management. To adapt to the development trend of the big data era and improve the classification effect of media images, this study introduces a dense connection refinement network (DRCN) in convolutional neural network image recognition, combined with attention mechanism, to fully utilize image features of different scales and improve the judgment accuracy of object detection based on increasing feature reuse times. The results indicate the consistent loss value of 0.08 for the DRCN-Attention model, while achieving a peak recall rate of 85% after 30 iterations, and a mean average precision exceeding 80%. The classification accuracy of ships reaches 82%, which is 8% higher than the support vector machine model. This indicates that the proposed media image classification method has high classification accuracy, provides a new technical reference for the field of computer vision, and has certain application value in the intelligent management of media images in the era of big data.


6. Research on Cooperative Early Warning Coastal Air Environment of Liaoning Based on Multi-Feature and Multi-Model Fusion
   Xiuli Yu, Peng Du
   Vol. 21, No. 3, pp. 296-307, Jun. 2025  
   https://doi.org/10.3745/JIPS.04.0350
   Details Full Text PubReader
   Keywords: Cooperative Early Warning, Ensemble learning, Multi-Feature Fusion, Multi-Model Fusion
   Show / Hide Abstract

   Enterprises in the coastal regions of China release abundant pollutants that have considerably deteriorated the air quality. To address this issue, an information fusion technology has been proposed herein for predicting coastal air quality in Liaoning Province. To this end, real-time data analysis of water, air, and soil pollutants emitted from diverse coastal enterprises was performed using a multi-model selection strategy for ensemble learning. This approach integrated meteorological information and considered the unique learning principles and observational disparities among various algorithms. The proposed approach explored the influence of collaborative early warning of multi-feature pollution source emissions on the surrounding environment. By combining the base learner and meta-learner in the multi-model fusion strategy, the ensemble model yielded better prediction results, particularly using strong learners at the primary level and linear mode at the secondary level. This optimal combination strategy was used to develop a collaborative monitoring and early warning model, which incorporated multi-feature data from water, air, and soil sources in the coastal environment. This multi-feature collaboration enhanced the prediction accuracy compared with that of single-feature models and further amplified the early warning capabilities enabled by multi-model fusion.


7. A Quantified Analysis of Urban Sprawl Based on High-Resolution Satellite Remote Sensing
   Dongmiao Zhao, Guangyi Zhang, Xiuhe Yuan, Chao Liu, Yansu Qi, Xingtian Wang
   Vol. 21, No. 3, pp. 308-317, Jun. 2025  
   https://doi.org/10.3745/JIPS.04.0351
   Details Full Text PubReader
   Keywords: Ecological Balance, GF-2, High-Resolution Satellite Images, Sprawl
   Show / Hide Abstract

   Under the circumstance of global sustainability, the expansion of urban has been paid serious attention by governments. To obtain the harmonious symbiosis with nature, decision-makers need a reasonable method to quantify and monitor the urban expansion. This paper analyzes a particular area through spatiotemporal identification and quantification by using high-resolution satellite images. The change trend and ratio of artificial construction areas in the research area from 2016 to 2023 are analyzed. Optical flow is chosen for visualization and analyzing the characteristics of urban expansion in the research area, directly expressing the quantity and direction of construction in the process of urban expansion, which cannot be reflected in the traditional image quantitative analysis. The results show that the urban is expanding to the coastline and causing irreversible damage to local natural environment. The proportion of vegetation coverage should be strictly controlled.


8. Anomaly Detection of Power Load Based on Robust PCA and Improved K-Means Clustering Algorithm
   Xinjian Zhao, Weiwei Miao, Song Zhang, Youjun Hu, Shi Chen
   Vol. 21, No. 3, pp. 318-327, Jun. 2025  
   https://doi.org/10.3745/JIPS.03.0205
   Details Full Text PubReader
   Keywords: Anomaly Detection, K-Means, feature extraction, Power Load, Robust Principal Component Analysis
   Show / Hide Abstract

   The interaction of power load information provides reliable data support for accessing user-side electrical energy storage devices and distributed renewable energy sources. However, owing to the large volume of interactive information and the numerous security threats faced during the interaction, anomaly detection has become one of the most challenging problems in smart grids. To address this issue, an anomaly detection method was developed that consists of three stages. First, feature extraction is performed based on the power load information. Then, a robust principal component analysis method is used for the preliminary classification of the extracted features. Finally, an improved K-means clustering algorithm is employed to refine the classification results into completely non-overlapping groups and detect anomalies from the classified data. Experimental results demonstrate that the proposed method can effectively and accurately detect anomalies from power load data.


9. Scalable Representation of Customer Purchase Preferences through Co-Purchase History
   Seonghyun Kim, Doyeon Kwak
   Vol. 21, No. 3, pp. 328-341, Jun. 2025  
   https://doi.org/10.3745/JIPS.04.0352
   Details Full Text PubReader
   Keywords: Co-purchase Recommendation, Customer Lifestyle, e-Commerce, Predictive Analysis, Purchase Preferences, retail, RFM
   Show / Hide Abstract

   In the competitive e-commerce landscape, accurately measuring customer preferences and effectively representing customer segments are essential for driving personalized marketing and product offerings. Current data-driven methods often rely on resource-intensive algorithms, and there is a need for a systematic and scalable framework for extracting product sets that represent specific purchasing preferences. This study proposes an unsupervised, efficient framework that leverages purchase history data to derive product sets that best represent known customer segments and product categories. Utilizing an item-based top-N recommendation technique, the proposed method tracks co-purchase histories and generates relevant novel segment variants, capturing hidden purchase preference attributes and delivering a more accurate depiction of customer behavior. Evaluation with real-world customer data from a Korean retail and e-commerce platform network substantiates the practical applicability of the suggested framework in forecasting the probability of purchasing target products, outperforming other prediction techniques. By adopting this scalable and readily implementable approach, businesses can effectively make well-informed decisions regarding product offerings, promotional campaigns, and personalized recommendations, ultimately improving customer engagement and sales.


10. Research on Maximum Power Point Tracking Algorithms for Photovoltaic Cells
    Shi-jin Xin, Xiao-feng Wang, Li-liang Jia, Han-rui Zhang, Bao-jie Wang, Jing-hua Li, Yan-ling Wang
    Vol. 21, No. 3, pp. 342-353, Jun. 2025  
    https://doi.org/10.3745/JIPS.04.0353
    Details Full Text PubReader
    Keywords: MATLAB Simulink, MPPT, Photovoltaic cell
    Show / Hide Abstract

    Owing to the influence of environmental changes, such as temperature and lighting, on the output power of photovoltaic cells, tracking the maximum power point of photovoltaic cells is of great interest. This paper provides a detailed introduction to the relevant theories of maximum power point tracking in photovoltaic cells. Subsequently, a simulation module is built using MATLAB Simulink to analyze the simulation results of tracking the maximum power point under sudden changes in temperature or lighting using disturbance observation, conductivity increment, and constant voltage methods. The advantages and disadvantages of these three methods were analyzed. The experimental results indicate that the constructed simulation algorithm module is correct, and the simulation results provide reference value for the selection and application range of maximum power point tracking algorithms in actual photovoltaic systems.