Dynamic clustering based on Minkowski similarity for web services aggregation

This research, introduces a new dynamic clustering method offering a new approach utilizing Minkowski Distance methods for calculating similarity of xml messages to effectively compress and aggregate them. The increase in Web services utilization has led to bottlenecks and congestion on network links with limited bandwidth. Furthermore, Simple Object Access Protocol (SOAP) is an eXtensible Markup Language (XML) based messaging system often utilized on the internet. It leads to interoperability by facilitating connection both users and their service providers across various platforms. The large amount and huge size of the SOAP messages being exchanged lead to congestion and bottlenecks. Aggregation tools for SOAP messages can effectively decrease the significant amount that traffic generated. This has shown a notable enhancement in performance. Enhancements can be made by using similarity methods. These techniques group together multiple SOAP messages that share a significant level of similarity. Present techniques utilizing grouping for aggregating XML messages have demonstrated efficiency and compression ratio limitations. Practically, the proposed model groups messages into clusters based on minimum distance, supporting Huffman (variable-length) and (fixed-length) encoding compressing for aggregating multiple compressed XML web messages into a single compact message. Generally, the suggested model’s performance has been evaluated through a comparison with K-Means, Principle Component Analysis (PCA) with K-Means, Hilbert, and fractal self-similarity clustering models. Minkowski distance clustering model has shown excellent performance, especially in all message sizes like small, medium, large, V.large. Technically, the model achieved superior average Compression Ratio and it has outperformed all other models.