Small Profits and Quick Returns: A Practical Social Welfare Maximizing Incentive Mechanisms for Deadline-sensitive Tasks in Crowdsourcing

This is python code for the paper work published in IEEE Internet of Things Journal. You can access to the paper through this link.

Prerequisites

• Language: Python
• Required Packages: numpy, pandas, matplotlib, munkres
• To install the required packages, type the following command:

1. Python 2

pip install numpy pandas matplotlib munkres

2. Python 3

pip3 install numpy pandas matplotlib munkres

Running the code

• Class 'task': task.py, Class 'provider' or 'worker': provider.py, and Class 'platform': platform.py
• Since some of simulations may take quite a long time, simulation results are attached in 'data' directory.

1. Visualization of the behavior of task

python3 task_figure.py

2. Visualization of the punctual behavior of worker

python3 provider_figure.py

3. Simulation for optimal task-worker matching

python3 optimal_run.py

4. Visualization of the results of 3)

python3 optimal_visualization.py

5. Simulation for one time competition between our work and the benchmark

python3 capacity_run.py

6. Visualization of the results from 5)

python3 capacity_visualization.py

7. Simulation for reselection process which consider the movement of workers and tasks between two platforms

python3 reselection.py

8. Visualization of the results from 7)

python3 reselection_visualization.py

9. Simulation for multiple reselection process which converges to balance point

python3 reselection_multiple_round.py

10. Visualization of the results from 9)

python3 reselection_multiple_round_visualization.py

11. Simulation for parameter setting

python3 parameter_run.py

12. Visualization of the results from 11)

python3 parameter_visualization.py