CrowdCog: A Cognitive Skill based System for Heterogeneous Task Assignment and Recommendation in Crowdsourcing
Source code for the prototype framework built for our study on using cognitive skills for heterogeneous task assignment and recommendation in crowdsourcing. If you use our work in your research, please cite the following paper published at CSCW 2020.
Danula Hettiachchi, Niels van Berkel, Vassilis Kostakos, and Jorge Goncalves. 2020. CrowdCog: A Cognitive Skill based System for Heterogeneous Task Assignment and Recommendation in Crowdsourcing. Proc. ACM Hum.-Comput. Interact. 4, CSCW2, Article 110 (October 2020), 22 pages. https://doi.org/10.1145/3415181
The current implementation includes five cognitive tests (Stroop, Flanker, Task Switching, N-Back and Self-ordered Pointing) and four crowdsourcing tasks (Classification, Counting, Transcription, and Sentiment Analysis). Please refer to the paper for more details including deployment conditions.
- MySQL
sudo apt-get install mysql-server - Venv
Create a new virtual environment at the project root
virtualenv -p python3 cognitive
Activate the virtual environment
source cognitive/bin/activate
Install dependencies
pip install django
pip install mysqlclient
Launch Django console
python manage.py shell
Build models for cognitive tests in django console
exec(open('workertasks/assignment/build_cognitive_model.py').read())
Create data in django console
exec(open('workertasks/initialize_data.py').read())
Check if the server is running
python manage.py runserver
Install the following
- Apache2
- Apache2 Mod WSGI
Create a new configuration named cognitive.conf at /apache2/sites-available with content from apache-site-cognitive.conf in project root
Enable the site
sudo a2ensite cognitive.conf
Reload apache config
sudo systemctl reload apache2
Please feel free to contact Danula Hettiachchi (contact@danulahettiachchi.com) for further questions.