
Anton Eklund
Graduate student @ Umeå University, Adlede AB, Umea, Sweden
bert
topic modeling
umap
hdbscan
human evaluation
stellar
1
presentations
1
number of views
SHORT BIO
I am a third year industrial Ph.D. student working at Adlede and associated with the Foundations of Language Processing group at Umeå University. My research interests revolve around document clustering and topic modeling. Particularly, I look at topic modeling done by clustering pretrained language model (PLM) embeddings. I have also developed a tool called STELLAR which is made for analysis and human evaluation of topic models. The current goal with my research is to investigate the possibilities of using clustering of PLM embeddings to model topic change over time, also called dynamic topics modeling.
Presentations

Topic Modeling by Clustering Language Model Embeddings: Human Validation on an Industry Dataset
Anton Eklund and 1 other author