Configurations

CAPRI contains a config.py file that provides adjustments and configurations for running experiments. One of the benefits of CAPRI is that the outcomes of the previously ran experiments and recommendation computations will be store in a defined directory. Thus, users can always follow-up and explore additional properties of the results. The parameters that can be set using this file includes:

dataDirectory

It contains the path to read data from.

outputsDir

It holds the path to store recommendation lists.

topK

Refers to Top-k items for evaluation (default: 10).

limitUsers

Limiting the number of users in dataset (default: -1)

listLimit

Limiting the length of recommendation lists (default: 10)

activeUsersPercentage

Calculating a list of defined percentages of users as “active users”

models

List of available models to be selected by users.

datasets

List of available datasets to be selected by users

fusions

List of available fusions to be selected by users

evaluationMetrics

List of available evaluation metrics to be selected by users