ProofWiki:Current events

If you have an idea for a current event add it to this page, or check out the community portal to see what's happening on $\mathsf{Pr} \infty \mathsf{fWiki}$.

See ProofWiki:Community Portal/Other Websites for the links to other websites, which has been moved from here.

AI Mashup Challenge

From: "Christoph LANGE" <ch.lange@jacobs-university.de>
Sent: Wednesday, November 30, 2011 4:57 PM
Subject: Call for Submissions: AI Mashup Challenge @ ESWC 2012; Deadline March 31

Dear all,

the AI mashup challenge saw math mashups in 2009 and 2010 – and in 2012?

Secondly, wikis are excellent platforms for prototyping and actually developing mashups – so how about some intelligent, wiki-based math mashups?

Cheers,

Christoph

------------------------------------------------------------
***** Call for Submissions and Papers *****
------------------------------------------------------------


AI Mashup Challenge 2012

of the

9th Extended Semantic Web Conference (ESWC)

May 27 - 31, 2012, Heraklion, Greece

Topics of interest

A mashup is a lightweight (web) application that offers new functionality by combining, aggregating and transforming resources and services available on the web.

The AI mashup challenge accepts and awards "intelligent" mashups that use AI technology, including but not restricted to

• machine learning and data mining
• machine vision
• natural language processing
• reasoning
• ontologies and the semantic web.

The emphasis is not on providing and consuming semantic markup, but rather on using intelligence to mashup these resources in a more powerful way.

Some examples:

• Information extraction or automatic text summarization to create a task-oriented overview mashup for mobile devices.
• Semantic Web technology and data sources adapting to user and task-specific configurations.
• Semantic background knowledge (such as ontologies, WordNet or Cyc) to improve search and content combination.
• Machine translation for mashups that cross language borders.
• Machine vision technology for novel ways of aggregating images, for instance mixing real and virtual environments.
• Intelligent agents taking over simple household planning tasks.
• Text-to-speech technology creating a voice mashup with intelligent and emotional intonation.
• The display of Pub Med articles on a map based on geographic entity detection referring to diseases or health centers.

You find more detail on the website of the 2011 challenge:

Awards

• € 1750 sponsored by Elsevier
• Speech outfit from Linguatec
• 10 O'Reilly e-books

The challenge tries to mediate between a grassroot bar-camp style and standard conference organization. This means for submitters:

• You announce your mashup as soon as you are ready, simply sending an email to the organizers (address below).
• The deadline is March 31, 2012.
• At a subpage of the mashup website provided by the organizers, you explain your work and refer to its URL.
• Your mashup stays at your URL and under your control. You can go on improving it.
• At review time (March 31, 2012), reviewers need a 5 page paper (LNCS format) that explains the mashup.
• The reviewers select the most interesting mashups for presentation and vote during the conference.
• Vote is public for all conference participants, but the reviewer quota makes up 40%.
• Be prepared to a give a brief talk and a demo during the conference.
• Awards will be handed over during the conference, and everybody will congratulate the winners!

Organizers

• Brigitte Endres-Niggemeyer, Hannover, Germany
• Krzysztof Janowicz, Santa Barbara, USA

Program Committee

• Luis M. Vilches Blázquez, Madrid, Spain
• Christoph Lange, University of Bremen & Jacobs University, Bremen, Germany
• Emilian Pascalau, University of Potsdam, Germany
• Giuseppe Di Fabbrizio, AT&T Labs, Florham Park NJ, USA
• Jevon Wright, Massey University, Palmerston North, NZ
• Aidan Hogan, DERI Galway, Ireland
• Alexandre Passant, DERI Galway, Ireland
• Emanuele Della Valle, Politecnico di Milano, Italy
• Tomi Kauppinen, Muenster, Germany
• Thorsten Liebig, Univ. of Ulm & derivo GmbH, Germany

Main Contact

Brigitte Endres-Niggemeyer
Heidegrün 36
30179 Hannover

brigitteen@googlemail.com
0511 8441690 (Festnetz)
015154726114 (mobil)
Skype: brigitte30179


Christoph Lange, Jacobs University Bremen

http://kwarc.info/clange, Skype duke4701


My up-to-date presentations: http://www.slideshare.net/langec

ITP 2011 Videos out now!

26-Sept-2011: http://www.cs.ru.nl/mwitp/

Laugh at one of your fave contributors fumble his way through half an hour of waffle. And there's other people there as well.

ITP 2011: Conference Report

It was noted how similar in intent, if not execution, $\mathsf{Pr} \infty \mathsf{fWiki}$ is to PlanetMath. This was seen as being a Good Thing, in that it can be used as the basis for considerable collaboration.

We briefly discussed the idea that $\mathsf{Pr} \infty \mathsf{fWiki}$ and PlanetMath might merge at some point, -- but at present the idea we're most interested in exploring is plentiful cross-linking to offer better mutual support. I note that this is starting to happen already - the (current) latest entry on PlanetMath responds to a question which is raised on $\mathsf{Pr} \infty \mathsf{fWiki}$; and there exists a template in $\mathsf{Pr} \infty \mathsf{fWiki}$ allowing direct linking to PlanetMath.

Joseph Corneli, one of the movers and shakers at PlanetMath, is doing a research project on aspects of math wikis (details at http://metameso.org/~joe/thesis-outline.html). During the course of discussions he and I had over the course of the weekend, the following suggestions and intentions were expressed:

Respect

$\mathsf{Pr} \infty \mathsf{fWiki}$ appears to be treated with considerable respect within the MathWiki community. It has aspects which appear to be unique and compelling, and it is doing something which no other site does, and by all accounts, it seems to be doing it well.

Internationalization

This has already been mentioned in $\mathsf{Pr} \infty \mathsf{fWiki}$, and is an outstanding question on the main page. The consensus at the time was that yes, this would be a good idea, but may be too much like hard work at the moment.

However, we might be able to make a start. Definitions, for example. It was suggested that on each definition page we might include the name for the concept being defined in whatever languages we can muster, or that are considered "important", e.g. French, German, Russian, Latin(?), Italian, and whatever other main languages have consistently been used for presentation of mathematical papers.

The example of "local field" being "corps local" in French (literally: "local body"), while the "natural language" translation of "field" into French translates back into English as "meadow" is a case in point. So using an ordinary dictionary There appears to be no on-line dictionary for the translation of mathematical terms. It would be good for one to be developed.

Making the proof structures more rigid

We are already most of the way there with this one. All proofs are written in simple sentences, the shorter the better, and simple "modus ponens" style compound statements, for example:

Let $A$ be a $B$.
Let $C$ be a $D$.
From the result "All $B$ are $D$", it follows that $A$ is a $C$.

(Yes I know the above does not logically hold, this is just an example of style.)

In more extreme cases the entire proof is done symbolically in an "eqn" template structure.

One big suggestion was that this may make it considerably easier for $\mathsf{Pr} \infty \mathsf{fWiki}$ to be interfaced with an automatic theorem prover (which was the main purpose of this conference, to bring together the threads of the research which has been done over the last 40 years or so). Already this structure is compatible with tools such as "Mizar" (I'll look up a link to it later) and research should be done to see how such an interface may be accomplished.

More to the point, as the structure is so rigid, it should be possible to further enhance the automatic structuring of our pages, for example by expanding the concept of the "eqn" template to encompass proof lines written in "natural language".

What all this means is that our "terse" house style would need to be rigidly enforced throughout. This is actually de facto as it stands, as every page written is near-enough immediately restructured as it comes in. So it is considered more than just an idiosyncrasy, it is being viewed as an essential part of $\mathsf{Pr} \infty \mathsf{fWiki}$'s philosophy by those who are coming to take it seriously as a valuable addition to the mathematics.

We are currently running with th GNU FDL licence. It has been suggested that we amend this to a Creative Commons license. There are many of these available to use, but two interesting ones are:

• CC0: Any material appearing here is considered to be free and available to be used by anybody.
• CC-By-SA: Allows any material to be used by anyone, as long as its source is cited. (This is the license that is used by PlanetMath, and by Wikipedia.)

I know nothing about the legal aspects of copyright policy (beyond expressing an opinion that in this context they get in the way of the spread of mathematical knowledge), but what I would encourage is that we adopt one whereby everything is as freely accessed as possible. However, if we were to change the policy from GNU FDL, we would need to consult all the contributors (or major contributors) to ask whether they would be okay with this change.

This may not be a particularly arduous thing to do, considering that most of the mathematics on this site has been around for so long that it's bound to be public domain anyway.

Watch this space, as there may well be suggestions for enhancements in this area.

Sources, and the Learning Experience

We had a discussion about how we learn mathematics. How do we "learn" mathematics? We don't know - if we did we'd bottle it and sell it.

But the usual concept is that we get a book and work through it from page one to page $n$, linearly. $\mathsf{Pr} \infty \mathsf{fWiki}$ / PlanetMath etc. have no such linearity. But its very lack of linear structure is in some ways a limitation.

It is noted that the source works from which the pages have been constructed are added at the bottom. It would not take too much effort to add a "next / prev" link to the "BookReference" template so as to provide a direct link to the next proof / definition in the book, for each book in question. Then the linearity of the book structure can be exploited, and a user may be able to follow the course of study for any given book.

Obviously the pages in question are more than just the content of the books, so any such path will provide a significantly richer learning experience than would be obtained by just following the book.

Exercises

Something else we've touched upon in the past. We may want to develop a standard technique for adding exercises to the various proofs and definitions. The thinking behind this is still seriously a work in progress.

Okay, that's some of the stuff we discussed. I'll get back to this in a while.

--prime mover 13:55, 29 August 2011 (CDT)