Word2Vec applied to Recommendation: Hyperparameters Matter

TL;DR

This paper investigates the impact of hyperparameter tuning on Word2Vec embeddings in recommendation systems, revealing that optimizing these parameters significantly enhances performance and differs from NLP settings.

Contribution

It systematically evaluates hyperparameters in recommendation contexts, showing their critical influence and the need for task-specific tuning, unlike standard NLP practices.

Findings

Optimizing hyperparameters improves recommendation performance by up to an order of magnitude.

Optimal hyperparameters differ significantly between NLP and recommendation tasks.

Certain hyperparameters like negative sampling distribution and window size are particularly impactful.

Abstract

Skip-gram with negative sampling, a popular variant of Word2vec originally designed and tuned to create word embeddings for Natural Language Processing, has been used to create item embeddings with successful applications in recommendation. While these fields do not share the same type of data, neither evaluate on the same tasks, recommendation applications tend to use the same already tuned hyperparameters values, even if optimal hyperparameters values are often known to be data and task dependent. We thus investigate the marginal importance of each hyperparameter in a recommendation setting through large hyperparameter grid searches on various datasets. Results reveal that optimizing neglected hyperparameters, namely negative sampling distribution, number of epochs, subsampling parameter and window-size, significantly improves performance on a recommendation task, and can increase it by an order of magnitude. Importantly, we find that optimal hyperparameters configurations for Natural Language Processing tasks and Recommendation tasks are noticeably different.