Published May 16, 2020
| Version v1
Conference paper
Open
A Comparison of Unsupervised Methods for Ad hoc Cross-Lingual Document Retrieval
Description
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.
Files
LREC_cross_lingual_news_linking.pdf
Files
(131.6 kB)
Name | Size | Download all |
---|---|---|
md5:f419cfd229d47c8a86a09e76aac3cdc9
|
131.6 kB | Preview Download |