E-VIEW-alation: Interactive visual evaluation of collocation candidate extraction

About E-VIEW-alation

The E-VIEW-alation approach is described in the paper

Evert, Stefan; Uhrig, Peter; Bartsch, Sabine; Proisl, Thomas (2017). E-VIEW-alation – a large-scale evaluation study of association measures for collocation identification. In Electronic lexicography in the 21st century. Proceedings of the eLex 2017 conference, Leiden, The Netherlands.

We are currently working on an open-source release of the underlying software package, which will be made available as soon as all the visualization wrinkles have been ironed out.

Evaluation results

eLex 2017: A large-scale evaluation study of association measures for collocation extraction
- Evert, Stefan; Uhrig, Peter; Bartsch, Sabine; Proisl, Thomas (2017). E-VIEW-alation – a large-scale evaluation study of association measures for collocation identification. In Electronic lexicography in the 21st century. Proceedings of the eLex 2017 conference, Leiden, The Netherlands. (presentation slides, video)
- gold standard: selected entries from the BBI and OCD2 collocation dictionaries (Bartsch224 set)
- parameters: 20 association measures, 13 corpora, 8 context sizes, 4 frequency thresholds
- E-VIEW-alation interactive viewer