The PeARS Recipe

We have put here a few pointers to papers and tutorials that provide some technical background to PeARS. A short system description was published at WWW2016 and is available here.

PeARS: a peer-to-peer agent for reciprocated search - PDF

Aurelie Herbelot

University of Trento, Center for Mind/Brain Sciences

We have done a first evaluation of the accuracy of the system, which was published at the *SEM conference and is downloadable here.

You and me in a vector space: Modelling individual speakers with distributional semantics - PDF

Aurelie Herbelot

University of Trento, Center for Mind/Brain Sciences

Behrang QasemiZadeh

DFG Collaborative Research Centre 991 Heinrich-Heine Universität, Düsseldorf

PeARS is powered by a technique called distributional semantics. A very simple introduction to distributional semantics is available here

What is distributional semantics - HTML

Aurelie Herbelot

University of Trento, Center for Mind/Brain Sciences

If you are interested in a more technical and comprehensive introduction, check out the papers by Turney and Pantel (2010) or Clark (2012).

From Frequency to Meaning: Vector Space Model of Semantics - PDF

Peter D. Turney

National Research Council Canada

Vector Space Models of lexical meaning - PDF

Stephen Clark

University of Cambridge, Computer Laboratory

If you are hooked... have a look at this rather long bibliography,

Distributional Sematics Bibliography - PDF

Aurelie Herbelot

University of Trento, Center for Mind/Brain Sciences

with many papers covering topics from formal meaning representations to multimodal semantics.