A Hybrid Machine Learning Model for Efficient XML Parsing

doi:10.1109/ACCESS.2024.3520706

2025-01-01

Muhammad Amjad Ali

Minhaj Ahmad Khan *

Raihan Ur Rasool

DOI：10.1109/ACCESS.2024.3520706

原文链接

原文求助

摘要

En 中文

The Extensible Markup Language (XML) files are extensively used for representing structured data on the web for file configuration, exchanging data between distinct applications, web development, and many other applications. Consequently, effective parsing techniques are necessary for XML files to enhance the performance of applications. The existing parsing techniques have their strengths and weaknesses affecting the performance of applications. Researchers point out that the selection of an efficient and appropriate parser is the most challenging issue regarding a particular condition. This paper proposes a framework XML Parsing Optimization using Hybrid Machine Learning (XPOHML) that makes use of Artificial Neural Network (ANN) and Support Vector Machine (SVM) machine learning techniques for efficient XML parsing. The newly developed framework performs analysis and prediction of different XML parsers using profiling, classification, performance evaluation, and finally generates code for efficient parsing. The XML profiling phase of the XPOHML framework generates a dataset by evaluating the performance of PXTG, SAX, StAX, DOM, and JDOM parsing models on separate cores by applying numerous file sizes. The Classification phase produces the classification model by applying ANN and SVM techniques to identify the appropriate parsing model. The performance evaluation phase of XPOHML assesses the performance of both parsing models through classification metrics (accuracy). Additionally, based on evaluation outcomes, the code generation phase produces an efficient parsing model of XML files. The newly designed and developed XPOHML framework has shown a meaningful improvement in the performance of parsing XML files.

Keyword:

XML

Support vector machines

Classification algorithms

Prediction algorithms

Machine learning

Codes

Machine learning algorithms

Predictive models

Performance evaluation

Training

Artificial intelligence