* Re: quantum of critic
2011-07-19 5:07 quantum of critic Joyal, André
@ 2011-07-19 16:28 ` Bob Coecke
0 siblings, 0 replies; 2+ messages in thread
From: Bob Coecke @ 2011-07-19 16:28 UTC (permalink / raw)
To: Joyal, André; +Cc: Timothy Porter, categories
Dear Andre, Tim,
> Dear Tim,
>
> I recently heard that research on quantum computing and quantum topos
> are presently generously supported in England.
>
> Is that right? André
> I believe they are not doing too badly but am not directly concerned
> with them so perhaps Jamie or Bob could comment.
>
> Tim
I assume that the question is about category-theory related quantum-ish
research.
Firstly, as several UK based category theory groups are already aware of,
we now have a network grant to support travel between UK category theory
related groups, more general travel, and events:
http://gow.epsrc.ac.uk/NGBOViewGrant.aspx?GrantRef=EP/I03596X/1
funded by EPSRC, the main British funding body, and coordinated by Samson
Abramsky, Andreas Doering, Jamie Vicary and myself. (I used to fund many
such events for the past five years with my personal EPSRC fellowship, but
that is now running out)
Concerning the Butterfield-Isham-Doering type topos approach to physics,
Steve Vickers has a decent grant from EPSRC, through their mathematics
panel (I think). In addition to this, Andreas Doering has PhD students in
the area funded by Oxford's Computing Laboratory. There is also funding
in the Netherlands for this subject. I am aware of two grants in Nijmegen
(Landsman, Jacobs, Moerdijk).
For the quantum computing related research that relies on monoidal
categories, there is quite a bit more funding, in part because of the
popularity of the topic of quantum computing. Both Samson Abramsky and I
have five year EPSRC fellowships that buy us out of faculty duties, a
number of postdocs also had EPSRC fellowships, we have a number EPSRC
grants in the area, and many students are funded by the department or
university. But we are also receiving/received substantial funding from
EU and US sources, e.g. Office of Naval Research that also sponsors the
MFPS meetings, and some charitable organizations. However, currently
a substantial part of the focus of our group's research is on foundations
of quantum mechanics, but we complement this with actual tool design
(which fundamentally draws from categorical logic):
http://sites.google.com/site/quantomatic/
An example of the "impact" that this research has generated, and which
immediately has been recognized by EPSRC with a five year "career
acceleration" fellowship for Mehrnoosh Sadrzadeh, is a spin-off from the
categorical quantum computing research to natural language processing
(NLP) technology. It started with a paper entitled "Mathematical
Foundations for a Compositional Distributional Model of Meaning":
http://arxiv.org/abs/1003.4394
which made it to the cover heading of New Scientist (this indeed counts
as "impact" with respect to REF):
http://www.cs.ox.ac.uk/people/bob.coecke/NewScientist.pdf
This work provides an algorithm that provides meanings of sentences given
meanings of words, relying on the common categorical structure of Lambek's
pregroups and the category of vector spaces (both are compact closed).
The inspiration came from the categorical quantum computing research:
"meanings of words are "teleported" within a sentence so that they can
interact". :)
Concerning "impact" at the concrete technological end,
Mehrnoosh Sadrzadeh and my student Ed Grefenstette produced a paper
entitled "Experimental Support for a Categorical Compositional
Distributional Model of Meaning" ("Categorical" as in Category Theory,
not as in categorial grammar, which is different of course) :
http://arxiv.org/abs/1106.4058
accepted for Empirical Methods in NLP, the main conference on the subject,
which shows that for certain NLP tasks the algorithm implemented in a
certain manner yields way better results than existing methods.
While of course one can hide that category-theoretic backbone, these
results could never have been obtained without, and the diagrammatic
language for monoidal categories is a particularly useful tool here.
Finally, it's probably well-known that general categorical semantics in
computer science has been a well-funded area in the UK for quite a while,
and there are many faculty positions in the area, in Edinburgh, Cambridge,
Oxford, Manchester, Bath, Leicester, ...
- best wishes, bob
[For admin and other information see: http://www.mta.ca/~cat-dist/ ]
^ permalink raw reply [flat|nested] 2+ messages in thread