Feature selection to enhance android malware detection using modified term frequency-inverse document frequency (MTF-IDF)

This research synthesizes an evaluation of feature selection algorithm by utilizing Term Frequency-Inverse Document Frequency (TF-IDF) as the main algorithm in Android malware detection. The TF-IDF algorithm is used to filter Android features filtered before detection process. However, IDF is unaware to the training class labels and gives incorrect weight value to some features. Therefore, the proposed approach that is Modified Term Frequency – Inverse Document Frequency (MTF-IDF) algorithm give more focus on both sample and features to give correct weight value to some features. The proposed algorithm considered features based on its level of importance where weight given based on number of features involved in the sample. The related best features in the sample are selected using weight and priority ranking process using K-means. This ensures that only important malware features are selected in the Android application sample. These experiments are conducted on a sample collected from DREBIN. Comparison between existing TF-IDF algorithm and MTF-IDF algorithm have been made under various conditions such as tested on different number of sample size, different number of features used and integration of different types of features. The results showed that feature selection using MTF-IDF can improve Android malware detection analysis. It was proven that MTF-IDF is an effective Android malware detection algorithm regardless of different kinds of features or sample sizes used. MTF-IDF algorithm also proved that it can give appropriate scaling for all features in analyzing Android malware detection.

Name dc:type.qualificationname

mphil

Level dc:type.qualificationlevel

masters

Grantor dc:publisher.institution

Universiti Tun Hussein Onn Malaysia

Year dc:date.issued

2019

Author dc:creator

• Mazlan, Nurul Hidayah

• QA76 Computer software

Language dc:language

en