Grant: Engage Grants Program
University of Waterloo
$23,800.00
Applying neural representations to semantic text search
EF 2011-12
Résumé
Finding relevant information within a large set of data is a difficult task, and one that is vital to our modern
information economy (e.g. Google). Current algorithms for doing this are based on high-speed computers
storing and sifting through vast amounts of data, and yet can miss information that most people could look at
and immediately see that it is relevant. The goal of this project is to evaluate computer algorithms that are
based on recent findings on how people perform this search task. If software can search through documents in a
manner that is more similar to human performance than current systems, then it can produce more realistic
recommendations as to what documents are relevant.
The core algorithms we will be using are Holographic Reduced Representations. These encode structured
information in a distributed manner, and provide a uniquely plausible account of how complex information is
stored in the brain (other accounts require too many neurons or peculiar connections between those neurons).
They allow us to encode both syntactic (structural) and semantic (meaningful) information in a compressed
format.
We will be evaluating search algorithms based on Holographic Reduced Representations in comparison to the
proprietary search algorithms used by DossierView. The intent is to discover whether these realistic human-like
representations can complement and improve on DossierView's existing search systems.
Programme
Engage Grants Program
Bénéficiaire
Emplacement
Ontario, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
418273-2011
Name nom
Eliasmith, Chris
Department département
Systems Design Engineering
Organization id
33
Institution établissement
University of Waterloo
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2011
Competition year année de concours
2011
Award amount
23800
Program id
EGP
Program nane en
Engage Grants Program
Program name fr
Programme de subventions d'engagement partenariat
Group en
Research partnerships programs
Group fr
Programmes de partenariat
Committee code
1553
Committee name en
Ontario Internal Decision Committee
Committee name fr
Comité de décision interne pour l'Ontario
Area of application code
802
Area of application group en
Information and communication services
Area of application group fr
Services d'information et de télécommunications
Area of application en
Information systems and technology
Area of application fr
Systèmes d'information et technologie de l'information
Research subject code
2803
Research subject group en
Artificial intelligence
Research subject group fr
Intelligence artificielle
Research subject en
Knowledge representation
Research subject fr
Répresentation des connaissances
Partie
1
Nb partie
1
Application title
Applying neural representations to semantic text search
Keyword
NSERC
Application summary
Finding relevant information within a large set of data is a difficult task, and one that is vital to our modern
information economy (e.g. Google). Current algorithms for doing this are based on high-speed computers
storing and sifting through vast amounts of data, and yet can miss information that most people could look at
and immediately see that it is relevant. The goal of this project is to evaluate computer algorithms that are
based on recent findings on how people perform this search task. If software can search through documents in a
manner that is more similar to human performance than current systems, then it can produce more realistic
recommendations as to what documents are relevant.
The core algorithms we will be using are Holographic Reduced Representations. These encode structured
information in a distributed manner, and provide a uniquely plausible account of how complex information is
stored in the brain (other accounts require too many neurons or peculiar connections between those neurons).
They allow us to encode both syntactic (structural) and semantic (meaningful) information in a compressed
format.
We will be evaluating search algorithms based on Holographic Reduced Representations in comparison to the
proprietary search algorithms used by DossierView. The intent is to discover whether these realistic human-like
representations can complement and improve on DossierView's existing search systems.