الفهرس 
Chapter 1 IntroductionOnly 14 pages are availabe for public view
 |  | from | 106 |  |  |
| | | from | 106 | | |
Abstract
Smartphones are quickly becoming necessities for many people. Smartphones are now used for more than just making and receiving phone calls. They can manage social networks, send emails, send money online, do online shopping, book tickets online, watch videos, do research, and a lot of other services. Unfortunately, they also serve as prime locations for malware to be spread by hackers. Because data movement has increased in our modern time, security is crucial. Malware causes unexpected behavior in smartphones such as strange charges on your phone bill, invasive adverts, contacts receiving strange messages, poor performance, appearance of new applications, abnormal data consumption and noticeable reduction in battery life. Android Operating System is an open-source operating system primarily designed for mobile devices such as smartphones and tablets. Android has since become the world’s most widely used mobile operating system. Android is a secure operating system, and Google uses a variety of security methods to safeguard users. However, there is a chance that malware will infect devices running Android, especially if users engage in dangerous conduct or download applications from dubious sources. Running computationally demanding antivirus algorithms on smartphones is hard because they require too much storage, processing, and communication overhead.
We present a comparative study of android mobile static analysis. Static analysis is used to classify malware android applications through meta data file of APK. Furthermore, we used TF-IDF and word2Vec feature extractors and investigate algorithms for static analysis, such as decision tree, naïve bayes, random Forest, K-nearest neighbor, XGB, MLP, support vector machine, logistic regression, adaboost, lasso regression, ride regression, ANN and extra trees. Two datasets, the small and large ”Drebin” datasets, are utilized in our research. Subsequently, we save the model obtained from the classifier that demonstrated
superior accuracy. This saved model is then employed in the detection of malware applications.
One answer to the malware problem affecting Android devices is the installation of a reputable antivirus app. An application for security or antivirus from a reliable vendor should be utilized. Your device can be scanned for malware, and ongoing security measures can be provided by these applications.
An Android antivirus application was developed, employing static analysis to detect whether any application contains malware. The antivirus utilizes a TF-IDF feature extractor and extra trees classifier. Two experiments were conducted:
In the first experiment, the antivirus is downloaded on Android devices, and its execution entirely occurs on the Android devices.
In the second experiment, the malware detection process runs on a server to implement the cloud computing concept. Cloud Computing involves using a remote server network instead of a local server or a personal smartphone to store, manage, and process data. Detection time was compared in each experiment.
The results of the comparative study for Android mobile static analysis algorithms indicate that TF-IDF performs better than Word2Vec feature extractor on the ”Drebin” dataset. Experiments on a small dataset reveal that multi-layer perceptron (MLP) yields the best overall accuracy of 98.84%, but it also takes the longest execution time, approximately 33.4 seconds. Conversely, results from a large dataset show that Extra trees achieve the best overall accuracy of 99.48%.
The results of the antivirus demonstrate that using cloud computing saves time and overhead. Cloud antivirus provides a fast and lightweight method to scan an Android application.