# A Method for Handling Negative Similarities in Explainable Graph Spectral Clustering of Text Documents -- Extended Version

arXiv: https://arxiv.org/abs/2504.12360 · HTML: https://mlsanity.com/paper/2504.12360/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 12 of its 31 references that are in the map.
- First posted on arXiv 16 Apr 2025.

## Builds on (most cited first, 12 of 12)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [A Survey of Transformers](https://mlsanity.com/paper/2106.04554/) (2021)
- [word2vec Parameter Learning Explained](https://mlsanity.com/paper/1411.2738/) (2014)
- [An Empirical Evaluation of doc2vec with Practical Insights into Document Embedding Generation](https://mlsanity.com/paper/1607.05368/) (2016)
- [A Survey of Signed Network Mining in Social Media](https://mlsanity.com/paper/1511.07569/) (2015)
- [On spectral partitioning of signed graphs](https://mlsanity.com/paper/1701.01394/) (2017)
- [Spectral Theory of Unsigned and Signed Graphs. Applications to Graph Clustering: a Survey](https://mlsanity.com/paper/1601.04692/) (2016)
- [On the Definiteness of the Weighted Laplacian and its Connection to Effective Resistance](https://mlsanity.com/paper/1408.2187/) (2014)
- [On Eigenvalues of Laplacian Matrix for a Class of Directed Signed Graphs](https://mlsanity.com/paper/1705.04406/) (2017)
- [Explainable Graph Spectral Clustering of Text Documents](https://mlsanity.com/paper/2308.00504/) (2023)
- [Edge-enhancing Filters with Negative Weights](https://mlsanity.com/paper/1509.02491/) (2015)
- [On Spectral Properties of Signed Laplacians with Connections to Eventual Positivity](https://mlsanity.com/paper/2009.03581/) (2020)

## Built on by (most cited first, 1 of 1)

- [Explainable Graph Spectral Clustering For GloVe-like Text Embeddings](https://mlsanity.com/paper/2508.14075/) (2025)

## Abstract (arXiv)

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.