Hello,
This is our first project, designed and developed by Tewfik Ghariani and Yosr Gouddi, students at ENSI.
It is a chrome plug-in that enables the analysis of comments associated to a facebook post, photo or video.
The extension is built using JavaScript. Concerning the server side, we used python as a programming language, as well as Tornado as a Python web framework since it's an asynchronous networking library
The analysis part is achieved using the a library called "TextBlob"
First, you have to install both "textblob" and "Tornado" before doing anything.
For "textblob" : pip install -U textblob For "tornado" : pip install tornado
How to use :
- The first step is to load the extension. Since it does not figure yet in the chrome store, the user must load the unpacked extension into any browser based on chromium.
Once the extension is loaded, the extension's icon will appear on the top right of the browser
- The user must run the python script via :
'python server.py'
-
Once loaded, the user has full control of the evaluation process. As a matter of fact, it is possible to enable the start of this process by simply clicking on 'Analyze Next Post'. This option is available inside the pop-up generated by clicking on the plug-in's icons. From now on, whenever the extension detects a Facebook post, photo or video, the analyze process will start.
-
After accessing a Facebook post page, the question enables the user to make a decision. In case the user refuses to start the process by submitting 'n', the operation is cancelled. Thus, the user must repeat the whole process by activating the extension. However, in case the user answers by inputting 'y', the mechanism starts immediately.
-
A simple notification is shown. At this point, the user should confirm this box while waiting for the results. Simultaneously, the data is being analyzed on the server's side in order to send back the results in a short amount of time.
The user has the privilege to visualize clear and obvious results. The results sent from the server are converted into meaningful colors. These colors help the user identify the nature of each comment to the extent of evaluating general feedback concerning the selected post
This was our very first project, and we are aware that it's not organized and it could have been a lot better. However, in order to improve ourselves, we would appreciate your advice and help.
Enjoy!