Lex2vec: making Explainable Word Embeddings via Lexical Resources

TL;DR

Lex2vec is a novel algorithm that enhances word embeddings by incorporating lexical resources and naming embedding dimensions with knowledge bases, improving interpretability.

Contribution

It introduces a method to inject lexical knowledge into embeddings and assign meaningful labels to dimensions, advancing explainability in word representations.

Findings

Extracted a set of informative, readable labels for embedding dimensions.

Achieved good coverage and interpretability of the embeddings.

Demonstrated the effectiveness of lexical resource integration in embeddings.

Abstract

In this technical report, we propose an algorithm, called Lex2vec that exploits lexical resources to inject information into word embeddings and name the embedding dimensions by means of knowledge bases. We evaluate the optimal parameters to extract a number of informative labels that is readable and has a good coverage for the embedding dimensions.