[Corpora-List] Geometrical representation of NL phrases for similarity comparison

Matthias Hartung mhartung at cit-ec.uni-bielefeld.de
Fri Oct 19 14:26:40 CEST 2018

Hi Alexander,

On 19.10.18 10:41, Alexander Osherenko wrote:
> I wonder if it is possible to represent NL phrases geometrically, for
> example, to compare their similarity. For example, the phrase "Hey man,
> that chick /is such a catch!/" and more formal "..., this girl is
> pretty!" should be represented geometrically nearby because they are
> semantically similar. 
> I am aware of LSA vectors that represent particular words and similarity
> could be evaluated as a distance between these word vectors in the LSA
> space. However, the LSA approach only works for individual words and no
> phrases and it is IMHO too numerical because it doesn't consider
> semantics of participating words.

you might want to consider the GEMS 2011 Shared Task about phrase similarity:


Best regards,


-- Dr. Matthias Hartung Semantic Computing Group Cognitive Interaction Technology Center of Excellence (CITEC) Bielefeld University

e-Mail: mhartung at cit-ec.uni-bielefeld.de Phone: +49 521 106 12223 Fax: +49 521 106 6560

Office: 2.310

Inspiration 1 / Zehlendorfer Damm 201 33619 Bielefeld Germany

More information about the Corpora mailing list