Systematic Review - Advances in Machine Learning Frameworks for Predicting Patent Infringements - Role of Algorithms and Data Processing

Aug 5, 2026

Upload by Asher

Patent infringement growth has increased the need for stronger, proactive intellectual property (IP) protection. This systematic review surveys advances in machine learning (ML) frameworks for predicting patent infringements, emphasizing algorithm performance, data balancing, and feature selection. Random Forest, Support Vector Machines (SVM), Logistic Regression, and hybrid ensembles are compared, with preprocessing methods such as SMOTE and Recursive Feature Elimination (RFE) highlighted for accuracy gains. Ethical issues, scalability, and bias mitigation are also discussed, ending with an integration roadmap for preventive IP risk management.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages9

Reading Time23 min

AccessFree to download

Introduction

Methodology

Methods

Systematic Review: Advances in Machine Learning Frameworks for Predicting Patent Infringements Kang Jin Ganga *, Ang Ling Weayb a,bMalaysia University of Science and Technology (MUST), Block B, Encorp Strand Garden Office, No. 12, Jalan PJU 5/5, Kota Damansara, 47810 Petaling Jaya, Selangor, Malaysia aEmail: [kang.jingang@phd.must.edu.my](kang.jingang@phd.must.edu.my) [b](bEmail: dr.ang@must.edu.my)[Email: dr.ang@must.edu.my](bEmail: dr.ang@must.edu.my) Abstract The rise of patent infringement cases has spurred the demand for innovative solutions in intellectual property (IP) management. This systematic review explores advancements in machine learning (ML) frameworks for predicting patent infringements, focusing on algorithm performance, data balancing, and feature selection. By evaluating Random Forest, Support Vector Machines (SVM), Logistic Regression, and hybrid ensemble models, we provide insights into their strengths and limitations. Key findings highlight the critical role of data preprocessing techniques, such as Synthetic Minority Oversampling Technique (SMOTE) and Recursive Feature Elimination (RFE), in improving model accuracy. Furthermore, ethical and practical considerations, including scalability and bias mitigation, are discussed. The review concludes by proposing a roadmap for integrating advanced ML techniques into proactive IP protection strategies. Keywords: Machine Learning (ML); Patent Infringement Prediction; Intellectual Property (IP) Management; Random Forest Algorithm; Hybrid Machine Learning Models. Received: 12/12/2024 Accepted: 2/5/2025 Published: 2/17/2025 * Corresponding author. 1. Introduction The protection of intellectual property (IP) is vital in today’s innovation-driven economy, as it serves as the foundation for fostering creativity, promoting technological progress, and ensuring competitive advantages for businesses and nations alike [1,2] . Among the various forms of IP, patents play a critical role by granting inventors exclusive rights to their innovations, thereby incentivizing further research and development [3,4] . However, as the volume and complexity of patent filings increase globally, these assets are becoming more susceptible to infringement, posing significant challenges for IP holders [5 ,6] . Traditional methods of IP protection, such as manual monitoring, litigation, and enforcement, are often reactive, resource-intensive, and insufficient to address the sophisticated and large-scale nature of modern infringement activities [7,8] . Consequently, the limitations of these conventional approaches highlight the urgent need for innovative solutions that can proactively safeguard intellectual assets. Machine learning (ML) has emerged as a transformative tool in this context, offering advanced analytical capabilities to predict and mitigate infringement risks with greater accuracy and efficiency [9, 10]. By leveraging techniques such as natural language processing, classification algorithms, and predictive modeling, ML frameworks can analyze extensive datasets to identify patterns indicative of potential infringements, enabling organizations to take preemptive actions [11, 12] . Recent advancements in ML, including hybrid ensemble models and deep learning techniques, further enhance the ability to handle complex, high-dimensional patent data [13, 14] . This review aims to synthesize the current state of ML frameworks for patent infringement prediction, evaluate the performance and advancements of various algorithms, and identify critical research gaps. By providing a comprehensive analysis, this review seeks to guide future developments in integrating ML into proactive intellectual property management systems. 2. Methodology 2.1 Methods A systematic search was conducted across peer-reviewed journals, conference proceedings, and grey literature to identify studies focusing on machine learning (ML) frameworks for patent infringement prediction. The

FAQs

What is the main goal of this systematic review?

To synthesize the current state of ML frameworks for predicting patent infringements, evaluate algorithm performance, and identify key research gaps for future proactive IP management.

Which ML approaches are evaluated in the review?

The review evaluates Random Forest, Support Vector Machines (SVM), Logistic Regression, and hybrid ensemble models, including discussion of strengths and limitations.

How do data preprocessing techniques affect prediction accuracy?

The review highlights that preprocessing methods such as SMOTE and Recursive Feature Elimination (RFE) improve model accuracy by addressing data imbalance and selecting more informative features.

Aug 5, 2026

Upload by Asher