Wallet App Credibility Analysis Based on App Content and User Reaction

WalletApp Credibility Analysis Based on App Content and User Reaction
Android PHP Projects
Abstract— Online reviews have become a valuable resource for decision making. In recent years, analysis of reviews has attracted significant attention. Individuals and organizations in making purchase and other organizational decisions are increasingly using online reviews. Positive reviews acclaim significant financial gains, fame and prestige for businesses in market. On the other end, this gives strong enticement to fraudulent to hypocrite the system by posting disingenuous reviews to promote or to vilify some target products and services, which is known as opinion spam. Therefore, we made this research effort as an initial impetus in the direction to identify the presence of fake reviews. We performed this experiment on Wallet Apps accessible through Google Play Store. Know More To do so, we rate applications based on features availability existence in Google App based on App content and user reactions. 
We computed four major scores- Description score, Positive Store, Negative Store and Review Tag Score and finally assigned a cumulative normalized score to all Wallet Apps. Further, a comparative analysis has been done between cumulative normalized score and stated Google App rating. Significant deviation indicates a strong probability that application has an opinion spam.Keywords—Opinion Spam; Google App; Wallet App; Fake Reviews; Information Retrieval

ApplIcAtIon credIbIlIty AnAlysIs process
 Opinion expressed in form of rating and reviews are considered as essential information for credibility analysis of an app. We have taken advantage of opinions expressed by users to analyze credibility of wallet app and this have been done in three phases - app data extraction phase, credibility score computation phase and credibility analysis phase as shown in Figure 2. App extraction data phase fetches all 192 Wallet-App content that contains eight features. 
We have used only four descriptive features, they are: app name, average rating, app reviews, and app description to perform our proposed empirical analysis process. Credibility score computation phase is the most critical and dominant portion of the research work in which we compute all complex feature-dependent credibility scores. 
The inputs considered four descriptive features for this phase and diverse computations performed in this phase are-ŸŸData Pre-processing ŸŸConstruction of Bag-of-Words (BOW1) for terms that are used in description of all appsŸŸApp Description score computation: it is computed using BOW1 and assign score to an app based on its word/term existence and its frequency (of a term) in app description on Google Play store. For example: if a word ‘secure’ is used 2 times in to the word ‘secure’ the app description, it assigns more weight to secure word and finally submission of all word score define app description scoreŸŸConstruction of Bag-of-Words (BOW2) for terms that are used in reviews of all apps

Comments

  1. Wow, that is quite informative. I like this article very much. The content was good. If any of the engineering students are looking for a Android Final Year Projects, I found this site and they are providing the best service to the engineering students regarding the projects Android Final Year Projects

    ReplyDelete

Post a Comment

Popular posts from this blog

ecriminal Identification Using Face Recognition Based Mobile App

Bridge between Investor and Business people

Mobile App for Soil with suitable farming agent and distributor location