opocaj92/Simulating-RS-in-Digital-Markets

Experimental setup for simulating the impact of recommender systems in digital markets

Simulating RS in Digital Markets

Experimental setup for simulating the impact of recommender systems in digital markets

Prerequisites

To be able to run the code provided in this repository it is necessary to install the extended t-recs library available at https://github.com/opocaj92/t-recs.

Please follow the instructions on the project page on how to install and configure the framework.

Usage

The provided code consists of three parts:

1. The running script main.py, used to run experiments,
2. The running and plotting functions, contained in utils.py,
3. The parameter files in params subfolders, that are used to provide the configuration parameters to the execution script.

A given param file (experiment configuration) can be run with:

python3 main.py -d subdir_of_param_file -p param_filename_without_py

Other available options are:

• -o output_subdir (default is a new folder with the name of the param file)
• -cbo to only execute content-based RS
• -hybrids to also execute hybrid RSs
• -plots to save plots on the tracked metrics
• -debug to save debug info (like the users/items values etc...)
• -more to track an additional set of metrics
• -skip used for big configs like paper_experiments/combined_all.py, it allows to skip configurations that will results in the same setting

Sweep Param Files

In order to simplify the execution of multiple parameter settings, we allowed for "sweep" param files, that contains a list of value for a given (or multiple) parameters. This will results in (sequentially) executing multiple configurations, one after the other, where the value of that parameter(s) takes different values in each execution.

As an example, if the parameter file is:

params = { a = 1, b = [1, 2], }

This will execute two different experiments, one where a = 1, b = 1 and the other where a = 1, b = 2.