Quantification Label for Mobile Apps

Abstract

Nowadays every product that we purchased has a tag informing its users how much an automobile or an appliance will cost to run and how much energy a food product will provide on consumption, for example. With the advent of Internet, purchasing apps become habitual, as Apps facilitate easy way of trading and generate money according to app online stores. Currently, Apps provide label contains a combination of developer info, memory size, category, operating system compatibility, user ratings and reviews, and privacy policy. However, the product label information is asymmetrical and it varies with the developer. Thus, it is essential to provide Apps with symmetrical metrics exhibiting the quality to the users, in similarity to food products labels. It is challenging to develop common standards explicit to users, as their behaviors are dynamic to draw insights from the history of tracked data. Thus, we propose a quantification label, which assesses the mobile/web App’s quality and facilitates the App developers and users to know how the selected App is featured, responsive, secured and energy-saving compared to similar Apps in the market. App undergoes a series of state changes during its execution and our proposed labeling scheme quantifies the changes in the enterprise/mobile network components during App execution. A set of parameters is selected to quantify the state changes, which are directly or indirectly influenced by the App, namely: degree and domain of connectivity, energy consumption and vulnerability, and a label frame comprises of feature: popularity, energy consumption, and security to reflect the quality of App. Further, the label utilizes the principal component analysis (PCA), which is a statistical technique, to compare the behavior of two or more Apps running simultaneously to categorize the prominent one impacting the network parameters. The experimental results in quantifying a real time traffic data of an internet based App validate the proposed label’s by revealing the key features to the users; moreover, PCA carried out on four sets of traffic data enables to select the prominent data and the results also reveal the possible embedding of the framework within an App for dynamic monitoring.