A Method for Handling Negative Similarities in Explainable Graph Spectral Clustering of Text Documents -- Extended Version
This addresses a specific technical issue in text clustering for researchers using modern embeddings, but it is incremental as it builds on existing solutions.
The paper tackles the problem of negative similarities in Graph Spectral Clustering (GSC) when using document embeddings like GloVe, which cause failures in normalized Laplacian-based methods. It shows that applying methods to cure similarity negativity improves accuracy for both combinatorial and normalized Laplacian-based GSC and enables explanation methods originally for Term Vector Space to work with GloVe embeddings.
This paper investigates the problem of Graph Spectral Clustering with negative similarities, resulting from document embeddings different from the traditional Term Vector Space (like doc2vec, GloVe, etc.). Solutions for combinatorial Laplacians and normalized Laplacians are discussed. An experimental investigation shows the advantages and disadvantages of 6 different solutions proposed in the literature and in this research. The research demonstrates that GloVe embeddings frequently cause failures of normalized Laplacian based GSC due to negative similarities. Furthermore, application of methods curing similarity negativity leads to accuracy improvement for both combinatorial and normalized Laplacian based GSC. It also leads to applicability for GloVe embeddings of explanation methods developed originally bythe authors for Term Vector Space embeddings.