waffiqaziz/ta-instafake

Fake Instagtram Account Detection using Support Vector Machine with PCA Feature Selection

InstaFake Dataset

Dataset of the Intagram Fake and Automated Account Detection paper

Installation

Install miniconda

https://conda.io/en/latest/miniconda.html

Setup a CONDA environment

We create a new vitrual environment named instafake.

conda create --name instafake python=3.6

Activating the virtual environment

If you are inside the virtual environment, your shell prompt should look like: (instafake) user@computer:~$ If that is not the case, you can enable the virtual environment using:

conda activate instafake 

To deactivate the virtual environment, use:

conda deactivate

Install required packages

To install the required packages, run the following command in your instafake virtual environment:

pip install -r requirements.txt

Import Datasets as Dataframes

To import the fake and automated datasets as pandas dataframes, simply define the dataset folder path data, and dataset version dataset_version and call import_data from utils:

from utils import import_data

dataset_path = "data"
dataset_version = "fake-v1.0"

fake_dataset = import_data(dataset_path, dataset_version)

dataset_path = "data"
dataset_version = "automated-v1.0"

automated_dataset = import_data(dataset_path, dataset_version)

Dataset Structures

The dataset contains of 2 set of json files with given features:

Fake Account Detection

1. user_media_count - Total number of posts, an account has.
2. user_follower_count - Total number of followers, an account has.
3. user_following_count - Total number of followings, an account has.
4. user_has_profil_pic - Whether an account has a profil picture, or not.
5. user_is_private - Whether an account is a private profile, or not.
6. user_biography_length - Number of characters present in account biography.
7. username_length - Number of characters present in account username.
8. username_digit_count - Number of digits present in account username.
9. is_fake - True, if account is a spam/fake account, False otherwise

Automated Account Detection

1. user_media_count - Total number of posts, an account has.
2. user_follower_count - Total number of followers, an account has.
3. user_following_count - Total number of followings, an account has.
4. user_has_highlight_reels - Whether an account has at least one highlight reel present, or not.
5. user_has_url - Whether an account has an url present in biography, or not.
6. user_biography_length - Number of characters present in account biography.
7. username_length - Number of characters present in account username.
8. username_digit_count - Number of digits present in account username.
9. media_comment_numbers - Total number of comments for a given media.
10. media_comments_are_disabled - Whether given media is closed for comments, or not.
11. media_has_location_info - Whether given media includes location, or not.
12. media_hashtag_numbers - Total number of hashtags, given media has.
13. media_upload_times - Media upload timastamps.
14. automated_behaviour - True, if account is an automated account, False otherwise

Dataset Metadata

The following table is necessary for this dataset to be indexed by search engines such as Google Dataset Search.

property	value
name	InstaFake Dataset: An Instagram fake and automated account detection dataset
alternateName	InstaFake Dataset
url	https://github.com/fcakyon/instafake-dataset
sameAs	https://github.com/fcakyon/instafake-dataset
sameAs	https://github.com/fcakyon/instafake-dataset
description	The InstaFake Dataset is comprised of anonymized Instagram user data collected by Fatih Cagatay Akyon and Esat Kalfaoglu over the second half of 2018. We’re releasing this dataset publicly to aid the research community in making advancements in machine learning based social media analysis.
provider	property value name Fatih C. Akyon and Esat Kalfaoglu sameAs https://scholar.google.com.tr/citations?user=RHGyDE0AAAAJ&hl=en
license	property value name Attribution-NonCommercial url https://creativecommons.org/licenses/by-nc/4.0//