0. Set up

Data

There are a few files that will be needed to run all the programs included here. Three of them including the training, validation and test datasets are included in this repository in the data folder. The collection of metadata was too large to include and can be found here. This set has already been cleaned.

Requirements

All the scripts provided here have been tested in python 3.6.9 in a linux environment. The requirements needed for all the scripts to run are detailed below:

• beautiful soup v4.9.1

• csv 1.0

• dash v1.14.0

• dash-bootstrap-components v0.10.5

• implicit v0.4.2

• lightfm v1.15

• numpy v1.18.1

• pandas v1.0.1

• requests v2.24.0

• scikit-learn v0.22.1

• scipy v1.4.1

• tqdm v4.48.0

All these requirements can be installed either with conda or with pip.

Warnings

In addition the file common_functions.py will be used by all the scripts. Take care that if you move the scripts from the current structure or move the common_functions.py file the scripts will not be able to find it and may fail. The dash application in particular should not be moved from its folder and should be executed from inside it due to its dependence on the presence of the files containing the information.