Gathering Security and Privacy Expectations of Users from Mobile Apps

Abstract

The scope of this project is to analyze the security concerns in mobile applications. There are some ways where we can check the security terms. One is the description of applications that are helpful while the user's behavior towards these applications. In this project, the reviews like user’s behavior towards application are analyzed with text analysis. Behavioral analysis is based upon text-like semantic analysis. Previously there are many techniques for text-based analysis problems. But in this report, the NLTK technique is used. NLTK has some benefits when working with text data. It supports the cleaning of the text and removing stop words in it, and the last one is NLTK correct those words which are incorrectly spelled and misplaced in a statement. The word which is not understandable is removed. And the stop words like dot, and, or, comma, it, is, him, her, are removed in this technique. And the cleaned corpus is created. After correcting the words, it will enable the algorithm to do further analysis on the dataset. The text-based dataset is solved by the one-hot encoding also. In this report, two datasets are used: reviews taken from the google play store, and the second one is taken from the apple app store. After the cleaning of words using NLTK and storing all words separately in Bag of words. This report finalized the supported words and show these words with the help of cloud pictures concerning security concerns. And the last is the comparison of both datasets where security words are chosen and displayed based on the user’s behavioral analysis techniques.