Conference paper Open Access
Zosa, Elaine; Granroth- Wilding, Mark; Pivovarova, Lidia
We address the problem of linking related documents across languages in a multilingual collection. We evaluate three diverse unsupervised methods to represent and compare documents: (1) multilingual topic model; (2) cross-lingual document embeddings; and (3) Wasserstein distance. We test the performance of these methods in retrieving news articles in Swedish that are known to be related to a given Finnish article. The results show that ensembles of the methods outperform the stand-alone methods, suggesting that they capture complementary characteristics of the documents.
|All versions||This version|
|Data volume||4.5 MB||4.5 MB|