What does this project do

django, sqlite3 based backend system for a google chrome extension that performs fake post injection in Facebook timeline and then conducts a survey on user experience.

SCREENSHOTS

Screenshots of the chrome extension that this backend system is a part of:

The front end exists in a different repository https://github.com/ishanguliani/fake-post-chrome-extension

This is how the fake chrome extension looks when installed

These are sample injected fake posts

Screenshots of the backend panel where you can upload your fake post models

Screenshots of the survey that the user takes after going through fake posts - this is all supported by the current python code

REQUIREMENT

To fire up the server have `python3` and `pipenv` installed.

open terminal and cd to the project root directory

start virtual environment

pipenv shell

install dependencies

pipenv install

set up database locally

python3 manage.py makemigrations 
python3 manage.py migrate

set up static files locally

python3 manage.py collectstatic

start the server

python3 manage.py runserver

Endpoints

Welcome screen: 127.0.0.1/
Admin panel: 127.0.0.1/admin
The survey: 127.0.0.1/surveyLinks/0/0

Generate report for link clicks

For all link user interactions happening while scrolling the extension, generate the following reports in real time -

Report showing user interaction summary

User	Total # links seen so far	Total # links clicked so far
Ishan Guliani	25	4
John Fletcher	12	0
Max Dean	50	17

Report showing user-link-click interaction

User	Link (actual link user would route to if clicked)	Clicked on	link_type	original link replaced	Clicked? (Y for yes)	# clicks on this link
Ishan Guliani	https://www.bbcnews.com	https://seng-research.com/jkhfdjk	genuine	-	Y	1
Ishan Guliani	https://www.newsusa.com	https://seng-research.com/aweoivl	fake	something.com	Y	2
Ishan Guliani	https://www.africanews.com	https://seng-research.com/kljadsnl	fake	something.com	NO	0
John Fletcher	https://www.youtube.com?video=dfcj897he1jfbkdvjyv9	https://seng-research.com/cxcsfbg	genuine	-	Y	1

TODO

Add a new section in the admin panel called Reports which will contain the above mentioned 2 reports namely -
- Real time click counts (showing how many links the user has seen or clicked in a session)
  - user - number of links seen - number of links clicked
- User link interaction report (showing what links the user was presented and what he did with them)
  - user - link type (genuine/fake) - link url - original link url (for links of type fake) - link short url - link_clicked (boolean) - number of times clicked
Add 3 new fields to the existing LinkModel
- is_clicked_event_from_ground_data (this field will be set to true when the user clicks on the link during the facebook session).
  NOTE: this is different from the existing field is_clicked in LinkModel which denotes the link clicks from the user's self reported survey data. As discussed, the idea here is to keep these two pieces of information different to be able to draw more conclusions on the disparity between the ground and reported data in the future which is a little misty for now.
- is_clicked_event_from_ground_data_time (this field will record the time of when the above event took place)

ishanguliani / fake-post-injector-backend Goto Github PK

fake-post-injector-backend's Introduction

What does this project do

SCREENSHOTS

Screenshots of the chrome extension that this backend system is a part of:

This is how the fake chrome extension looks when installed

These are sample injected fake posts

Screenshots of the backend panel where you can upload your fake post models

Screenshots of the survey that the user takes after going through fake posts - this is all supported by the current python code

REQUIREMENT

To fire up the server have python3 and pipenv installed.

open terminal and cd to the project root directory

start virtual environment

install dependencies

set up database locally

set up static files locally

start the server

Endpoints

fake-post-injector-backend's People

Contributors

Stargazers

Watchers

fake-post-injector-backend's Issues

Report showing user interaction summary

Report showing user-link-click interaction

TODO

Recommend Projects

Recommend Topics

Recommend Org

To fire up the server have `python3` and `pipenv` installed.