A Novel Malware Analysis Framework for Malware Detection and Classification using Machine Learning Approach

Sethi, Kamalakanta; Chaudhary, Shankar Kumar; Tripathy, Bata Krishan; Bera, Padmalochan

doi:10.1145/3154273.3154326

Cited by 33 publications

(23 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Neuro-Fuzzy is a complex model that combines fuzzy logic with neural networks to generate rules which are similar to our rules with the help of artificial neural networks as shown in figure 2. A perceptron is a basic unit in the neural network that computes to detect features in the inputted data [14]. Distance-Based Methods are based on predefined distance measurement used for classification.…”

Section: Methods Algorithmsmentioning

confidence: 99%

Emerging Machine Learning Techniques in Malware Detection and Analysis: A Comparative Analysis

Chintha¹,

Kumar²

2020

IJAR

View full text Add to dashboard Cite

New types of malware with unique characteristics are being created daily in legion. This exponential increase in malwareis creating a threat to the internet. From the past decade, various techniques of malware analysis and malware detection have been developed to prevent the efficacy of malware. However, due to the fast-growing numbers and complexities in malware, it is getting difficult to detect and analyze the malware manually. Because of the inefficiency in manual malware analysis, automated malware detection and analysis would be a better solution. Thus, malware analysis supported by machine learning became a required part of malware analysis. The automation used in learning patterns in malware can help in efficiently identifying the complexities. Malware Analysis with help the Machine learning would be more efficacious in terms of automation and memory usage. In this paper, we conducted a review of emerging various ML (Machine Learning) strategies used so far, in the field of malware analysis, to give a comprehensive view of the existing processes. We systemized them on various aspects like their objectives, machine learning algorithms used, information about the malware, etc. We also highlighted the existing problems in this particular field of study and tried to find multiple ways in which advancements can happen concerning the current trends being used.

show abstract

Section: Methods Algorithmsmentioning

confidence: 99%

Emerging Machine Learning Techniques in Malware Detection and Analysis: A Comparative Analysis

Chintha¹,

Kumar²

2020

IJAR

View full text Add to dashboard Cite

show abstract

“…Da mesma forma, [Ferrand 2015] demonstra algumas técnicas de detecc ¸ão de sandbox, mais especificamente voltadas para o Cuckoo Sandbox, contendo diversos códigos sobre a implementac ¸ão das mesmas e também possíveis contramedidas para tais técnicas. [Sethi et al 2018] propõem um framework para detectar e classificar diferentes arquivos como benignos ou maliciosos usando um classificador de dois níveis. Sua soluc ¸ão usa Cuckoo Sandbox para gerar um relatório de análise estática e dinâmica, executando os arquivos de amostra no ambiente virtual, bem como um módulo de extrac ¸ão de atributos que usa os relatórios gerados pelo Cuckoo Sandbox e emprega Weka para desenvolver modelos de aprendizado de máquina.…”

Section: Trabalhos Relacionadosunclassified

Sandbox as a Service: automatizando a configuracão do Cuckoo Sandbox e a geração de dados para análise de malware

Figueiredo

Cattelan

Miani

2022

Anais Estendidos Do XXII Simpósio Brasileiro De Segurança Da Informação E De Sistemas Computacionais (SBSeg Estendido 2022)

View full text Add to dashboard Cite

Este trabalho busca criar uma plataforma unificada de análise dinâmica e estática de malware, na qual diversos pesquisadores possam gerar conjuntos de dados para utilização em suas pesquisas. Propõe-se que a geração de tais conjuntos seja flexível o suficiente para abranger as diversas áreas de pesquisa de malware, permitindo escolher entre os campos obtidos a partir da fase de análise, que poderão ser incluídos ou não no conjunto de dados resultante. Através de extensões nas funcionalidades da plataforma Cuckoo Sandbox, foram implementadas novas funcionalidades, permitindo o download em lote dos relatórios gerados em formato CSV, mais adequado para algoritmos de aprendizado de máquina.

show abstract

“…The paper tells the process of taint assisted malware analysis and enhances the malware identification steps [7]. Kamla Kant Sethi et al gives a framework of malware analysis for classifying the malware with identification of the malware by using Sandboxing Tools and Machine level learning tools for extracting exact information about malicious software [8]. Li Li et al portray a systematic review on the need for static analysis in malware detection but the methodology that is described in the paper uses the automated tools for static code analysis of malware.…”

Section: Why Malware Analysis?mentioning

confidence: 99%

A Framework for Producing Effective and Efficient Secure Code through Malware Analysis

Pandey¹,

Tripathi²,

Alenezi³

et al. 2020

IJACSA

View full text Add to dashboard Cite

Malware attacks are creating huge inconveniences for organizations and security experts. Due to insecure web applications, small businesses and personal systems are the most vulnerable targets of malware attacks. In the wake of this burgeoning cyber security breach, this article propositions a framework for a complete malware analysis process including dynamic analysis, static analysis, and reverse engineering process. Further, the article provides an approach of malicious code identification, mitigation, and management through a hybrid process of malware analysis, priority-based vulnerability mitigation process and various source code management approaches. The framework delivers a combined package of identification, mitigation and management that simplifies the process of malicious code handling. The proposed framework also gives a solution for reused codes in software industry. Successful implementation of the framework will make the code more robust in the face of unexpected behavior and deliver a revolutionary stage wise process for malicious code handling in software industry.

show abstract

A Novel Malware Analysis Framework for Malware Detection and Classification using Machine Learning Approach

Cited by 33 publications

References 3 publications

Emerging Machine Learning Techniques in Malware Detection and Analysis: A Comparative Analysis

Emerging Machine Learning Techniques in Malware Detection and Analysis: A Comparative Analysis

Sandbox as a Service: automatizando a configuracão do Cuckoo Sandbox e a geração de dados para análise de malware

A Framework for Producing Effective and Efficient Secure Code through Malware Analysis

Contact Info

Product

Resources

About