Random

Introduction

When discussing recommendation systems, accuracy is often regarded as the most crucial metric. However, besides accuracy, several other key metrics can evaluate the effectiveness of a recommendation system, such as diversity, coverage, and efficiency. In this context, the random recommendation algorithm is a valuable baseline. In terms of implementation, for a given user and item, the random recommendation algorithm provides a random rating.

Running with RecBole

A Running Example:

Write the following code to a python file, such as run.py

from recbole.quick_start import run_recbole

run_recbole(model='Random', dataset='ml-100k')

And then:

python run.py

If you want to change parameters, dataset or evaluation settings, take a look at

• Config Introduction

• Data Module Introduction

• Training & Evaluation Introduction

• Usage