This is the code for the paper titled

Spotting Collective Behaviour of Online Frauds in Customer Reviews. Sarthika Dhawan*, Siva Charan Reddy Gangireddy, Shiv Kumar, Tanmoy Chakraborty

accepted at 28th International Joint Conference on Artificial Intelligence.

• Python To install the dependencies used in the code, you can use the requirements.txt file as follows -

pip install -r requirements.txt

Run the detection.py followed by refine_groups.py as follows -

python detection.py

The agruments it takes are (All are mandatory):

• --metadata: Path to metadata for the particular dataset.
• --rc: Path to review content for the particular dataset.
• --dg: Path to save the groups detected (json format).

python refine_groups.py

The agruments it takes are (All are mandatory):

• --metadata: Path to metadata for the particular dataset.
• --rc: Path to review content for the particular dataset.
• --groups: Path to groups generated by detection.py.
• --outputgroups: Path to save the output groups (json format).

This will generate fraud reviewer groups for the particular dataset.

Run the ranking.py as follows -

python ranking.py

The agruments it takes are (All are mandatory):

• --groups: Path to groups generated by refine_groups.py.
• --ef: Path to reviewer embeddings.
• --rankedgroups: Ranked group IDs (txt format, line separated IDs).

This will rank fraud reviewer groups for the particular dataset.
Provide appropriate paths for data files and parameters.

If you face any problem in running this code, you can contact us at sarthika15170[at]iiitd[dot]ac[dot]in.

For copyright (c) Sarthika Dhawan, Siva Charan Reddy Gangireddy, Shiv Kumar, Tanmoy Chakraborty

For license information, see LICENSE or http://mit-license.org