International Journal of Web and Grid Services (4 papers in press)

Regular Issues

• Bi-Phase LSTM: A LSTM-Based Autoencoder Architecture for Dynamic Social Network Prediction  
  by Hui Lin, Yi-Cheng Chen 
  Abstract: In recent years, social networks have grown in popularity, with most people actively engaging on these platforms. These networks hold valuable insights into users’ values and interests, allowing us to analyse relationships between connected individuals and even predict potential friendships. However, social networks are dynamic, and their structure evolves over time. To account for this, we employed a dual approach using a bi-phase LSTM autoencoder and a bi-phase LSTM predictor. These tools capture the changing characteristics of social networks and predict future graph structures. We rigorously tested our model on three datasets and compared its performance with other models. The bi-phase LSTM consistently delivered strong results across all datasets. Additionally, the model’s hyperparameters were fine-tuned to improve predictive accuracy, demonstrating its reliability in forecasting the evolution of social network structures.
  Keywords: feature extraction; autoencoder; decoder; long short-term memory; dynamic social network.
  DOI: 10.1504/IJWGS.2025.10068502
   
  
• Adaptive Proportional-Derivative Adjuster for Notch Filtering in Wearable ECG Monitoring  
  by Chao-Ting Chu, Jiawei Chang 
  Abstract: This paper presents an adaptive proportional-derivative adjuster (APDA) integrated into a notch filter for processing ECG signals in wearable smart clothing. The APDA filter dynamically adjusts parameters such as gain and cutoff frequency in real time to mitigate environmental interference, including 60 Hz noise and electromyographic signals common in workplaces and fitness centres. Unlike traditional fixed-parameter filters, this adaptive method minimises noise and preserves ECG clarity under varying conditions. A notable contribution of this study is optimising the APDA algorithm for resource-constrained microcontrollers, reducing both memory requirements and processing power. This streamlined computation lowers complexity and extends battery life, making it suitable for continuous, long-term monitoring. Experimental results from a smart clothing prototype validate the filters effectiveness in suppressing noise while maintaining signal integrity, offering an energy-efficient and cost-effective solution for wearable healthcare applications. This advancement addresses the limitations of conventional designs and shows promising potential for future clinical applications.
  Keywords: adaptive filtering; proportional-derivative control; wearable ECG monitoring; real-time noise suppression; sustainable healthcare technology.
  DOI: 10.1504/IJWGS.2025.10069918
   
  
• Web-Based Traditional Craft Kansei Retrieval Method using a Machine Learning Model and Feature Extraction  
  by Naho Kuriya, Kaisei Komoto, Tomoyuki Ishida 
  Abstract: In a previous study, we conducted a Kansei analysis of traditional crafts using the semantic differential method and correlated visual features with physical features extracted via visual pattern image coding (VPIC) from images of craftwork. However, associating the VPIC features with visual attributes is labour-intensive. Therefore, this study proposes a new Kansei retrieval method for traditional crafts. This method utilizes a machine learning model trained on the previous study's data to predict Kansei words based on features extracted from colour histograms and colour moments of craft images. To evaluate the proposed method, we conducted a questionnaire survey with 52 university students. Approximately 90% of the participants responded positively to the operability, visibility, relevance, and effectiveness of the entire system, and the functionality of the Kansei word retrieval function, while approximately 20% responded negatively to the functionality of the Kansei word prediction function.
  Keywords: Web application; Kansei retrieval; Kansei prediction; machine learning model; feature extraction.
  DOI: 10.1504/IJWGS.2025.10070264
   
  
• Improved Simplified Swarm Optimisation for Bipartite Graph Convolutional Network  
  by Zhenyao Liu, Wei-Chang Yeh 
  Abstract: Bipartite graphs have been widely applied in data mining to represent data relationships, such as in e-commerce recommendation systems. Graph neural networks (GNNs), with their powerful ability to process structured data and explore higher-order information, have become the state-of-the-art method for recommendation problems. Recommendation systems increasingly rely on graph structures to represent relationships between users and items, like user click behaviours and purchase records. Through graph convolutional networks (GCNs), these structures capture connections between users and items, integrating structural information (e.g., user-item links) with node features (e.g., user preferences and item attributes) for accurate recommendations. This study combines improved simplified swarm optimisation (iSSO) with bipartite graph convolutional networks and eye-tracking technology to explore user preference behaviour, called iSSO-BGCN. We construct a node-feature bipartite graph, using iSSOs optimisation capabilities and natural gradient descent to train the model. Trials validate its ability to deliver precise recommendations.
  Keywords: Graph Neural Networks; Graph Convolutional Networks; Improved Simplified Swarm Optimization; Bipartite Graph; Recommendation System.
  DOI: 10.1504/IJWGS.2025.10070494