Moghaddam

Friendship Preference: Scalable and Robust Category of Features for Social Bot Detection

---

• authors: Samaneh Hosseini Moghaddam, Maghsoud Abbaspour

• link: https://ieeexplore.ieee.org/abstract/document/9735340

• file structure:

├── process.py # convert raw dataset into standard format
└── train.py # train model on every dataset

• implement details: “Favorites count” is discarded since required information is not included in datasets.

How to reproduce:

1. process data, get the features and specify the dataset by running;

   python process.py --datasets "dataset name"

   there will be a .npy file, which is the processed feature

2. train random forest model and specify the dataset by running:

   python train.py --datasets "dataset name" > result.txt

   the final result will be saved into result.txt

Result:

random seed: 100, 200, 300, 400, 500

dataset		acc	precison	recall	f1
cresci-2015	mean	0.7361	0.9833	0.5923	0.7393
cresci-2015	std	0.0016	0.0026	0.0032	0.0021
Twibot-20	mean	0.7405	0.7229	0.8438	0.7787
Twibot-20	std	0.0080	0.0067	0.0103	0.0071
Twibot-22	mean	0.7378	0.6761	0.2102	0.3207
Twibot-22	std	0.0001	0.0010	0.0007	0.0003

baseline	acc on Twibot-22	f1 on Twibot-22	type	tags
Moghaddam et al.	0.7378	0.3207	F G	random forest