Landing page for the web application
We just have to get the url of the app, currently we will be considering a top rated app from google play store
We just have to paste the url and click on predict
And it will show you the result
We also get results to be "cautious" for probably average app, with poor reviews. (Not disclosed so that we don't get sued :) )
Similarly we can get results for apps from apple store
You know the drill !
It's interesting to see how some people completely ignore to research about the app before downloading and are not aware of malware, or poor interface they might face using a particular app. Hence apart from stars( which can be manipulated a lot easier than the reviews), we have made an application that will consider the reviews of users and give you a verdict hence after which the user can make a choice.
What it does
The application gives you the main verdict:
The verdict is pretty intuitive depending on the reviews you will get one of these 3 verdicts.
How we built it
We used Flask to build the application with simple UI. To get the reviews we have scrapped from the URL provided using google-play-scraper and apple-store-scraper and preprocessed it using standard preprocessing techniques. We have used semantic analysis to get the polarity of the reviews, to be specific we have used the dictionary method to do so.
Challenges we ran into
Running on various apps and getting a threshold for the verdict was quite challenging, and definitely can be improved.
Accomplishments that we're proud of
We were able to build a smooth-running web page that was really satisfying. :)
What's next for SafeApps
We do have a couple of things that can be improved, starting with the threshold. Testing on more apps would definitely improve the accuracy, along with this we can have a BiLSTM model to give us the polarity of the review too. We can also improve on the web-page and its design.