<p>Gestern abend haben ich in Begleitung einer weiteren Person die <a href="http://www.phatraake.de/index.php?sitemap">Phatcafer Laboratorien</a> besucht. Eine sehr interessante und vorallem auch sehr durch die musikalische Ader des <a href="http://www.phatraake.de/content/images/1_469_UpNpbD8HA3-190x.jpg">Herrn Becker</a> gepr�gte Umgebung.<br />Kreativ waren wir in dem Sinne, als dass wir ein gemeinsames mindset entwickelt haben. Also auf in eine bessere Zukunft… #B4mad.Net fully supports phatcafer lab.</p>
<<blockquote><p>The developer(s) of every Web 2.0 app/service should seriously consider exposing their data with a SPARQL query service.</p></blockquote><p>- <a href="http://www.oreillynet.com/pub/wlg/8609">Kendall Clark</a></p>
<<p>The <a href="http://business2.blogs.com/business2blog/2005/11/google_base_is_.html">Business 2.0</a> blog made the mistake of inserting Paul Ford’s little Googlebot image (from the <a href="http://www.ftrain.com/google_takes_all.html">Google/SemWeb piece</a>) into their content, using the original <code>img src</code> but without any attribution. Soon after, the article on Google Base at <a href="http://business2.blogs.com/business2blog/2005/11/google_base_is_.html">B2Day</a> played host to a bunch of rather <a href="http://www.ftrain.com/gooooooglebase.html">rude stoats</a>. Thus <a href="http://www.amazon.com/exec/obidos/tg/detail/-/0452286638/">Paul’s new novel</a> goes on my Xmas list ;-)</p>
<p>Some worth noting economy news as of Dec. 1, 2005:<ul><li>Personal saving rate improved to negative 0.7% from negative 0.8%. The record low as in August at negative 2.2%</li><p><li>Personal incoming increased 0.4% in Oct.</li><p><li>Consumer prices rose 0.1% in Oct, which is lower than 0.9% in Sept.</li></ul><p>It’s good to see that personal incoming has increased while the prices of our daily goods haven’t increase too much. Increasing personal incoming could also less the chance of seeing depreciation in real estate values. There are signs that people begin to save some of their incoming. Nevertheless, to me negative saving is unacceptable.</p><p>Source: <a href="http://www.marketwatch.com/news/story.asp?siteid=mktw&guid=%7BE64D8B24-3500-4C6F-8F5A-8C5372236994%7D&dist=bnb" title="Incomes, spending rise modestly">Incomes, spending rise modestly</a></p>
<p>I enjoyed the <a href="http://www.tomrafteryit.net/cork-geek-dinner/">Cork Geek Dinner with Robert Scoble</a> and all the other Irish geeks last night.</p><p>Again, it was good to meet not only the guest of honour but the other friends and acquaintances: Ed, Piaras, Damien, James, Bernie, Tom, Antoin, Fergal and Sven of Blogwise who came over from the UK.</p>
<p>Unofficial Fremont facts:<ul><li><a href="http://news.yahoo.com/s/cmp/20051201/tc_cmp/174403132" title="Microsoft Readies Online Classifieds Service">It took a 5 people team (including the manager is 6) to development Fremont</a></li><li>Fremont is a copycat of Google Base, not an original idea by Microsoft — I think. <u>Justification</u>: <a href="http://slashdot.org/article.pl?sid=05/10/25/2042238&tid=217" title=" Google Developing Database Service">Google Base was first announced on Slashdot on Oct. 25, 2005</a>. It’s quite possible that Fremont was developed under the pressure of the Microsoft management. Using <a href="http://en.wikipedia.org/wiki/Extreme_Programming" title="Extreme Programming">Extreme Programming</a>, I can believe how 5 experienced developers can build something like Google Base.</li><p></ul>
<p>Today I finished packaging up some of the Java Code I did. I released a <a href="http://b4mad.net/2005/12/01/WITW-webservice.jar"><code>WITW-webservice.jar</code></a> which contains some classes to use with the <a href="http://norman.walsh.name/2005/02/is/goern">Where in the World (WITW)</a> web service of <a href="http://norman.walsh.name/">Norman Walsh</a>. </p><p>The <code>.jar</code> contains the full source but no documentation jet…</p><p><em>TODO</em>: write documentation.</p>
<<p>Dan Connolly <a href="http://dig.csail.mit.edu/breadcrumbs/node/29">writes</a>:</p><blockquote><p>…the Department of Justice and the Department of Homeland Security announced the result of an XML collaboration - version 0.1 of the <a href="http://niem.gov/niem01.php">National Information Exchange Model (NIEM)</a> which will be used for law enforcement, emergency management, etc. communities and the parties who exchange information with them.</p><p>I hope to check it out. Better yet… I hope somebody else checks it out and writes a GRDDL transformation.</p></blockquote><p><a href="http://www.w3.org/2004/01/rdxh/spec"><abbr title="Gleaning Resource Descriptions from Dialects of Languages">GRDDL</abbr></a> uses XSLT to allow translation of existing XML docs into RDF/XML, effectively making them transparently visible on the <a href="http://www.w3.org/2001/sw/">Semantic Web</a>. </p><p>I couldn’t resist having a quick look. The zip file they provide contains 54 XML Schemas, and an Excel spreadsheet.The material in the spreadsheet is very structural, e.g.<br /><code><br />j:ActivityType extends c:ActivityType A structure that describes details about an activity or process that occurs.<br /></code><br />It seems like it might be relatively straightforward to mostly machine-process it into an RDF/OWL ontology, maybe using those <a href="http://jroller.com/page/rickard/20051030">abusive</a> tricks.</p><p>The XSD stuff, well, it’s XSD. But that could make a reasonable starting point for the XSLT. Sun have a <a href="http://www.sun.com/software/xml/developers/relaxngconverter/">RELAX NG Converter</a> which can eat XML Schema. A complication with the NIEM bundle is that it’s full of cross-references, looks like to do the conversions you’d need to dump the subdirectories in the package at the root of the filesystem. Then there’s another Java tool, <a href="http://www.relaxer.org/">Relaxer</a>, that can take Relax NG schemas as input and auto-generate identity XSLT transforms. The result isn’t wonderful, I seem to remember having to do search and replaces around the namespaces. But it does make a handy kind of template, and if the target structures aren’t that different from the source a lot of the work’s done.</p><p>A related tool is <a href="http://www.thaiopensource.com/relaxng/trang-manual.html">Trang</a>, which can generate Relax NG schemas from XML instance documents. </p><p>As Mr.C suggests, this job would make a nice student project, there’s bound to be some thinking & a fair bit of manual work in getting the ontologies and transformations right. But anyone does want to make potentially a lot of data more widely usable (and collect a few Connolly appreciation points), those little tools might help.</p>
<<p>In parallel with all the developments around <a href="http://microformats.org">microformats</a>, the <a href="http://structuredblogging.org/">Structured Blogging</a> folks have been taking a slightly different strategy (XML embedded in <code><script></code> elements) to broadly the same ends: getting data on the Web with the help of existing HTML-oriented tools. </p><p>In this <a href="http://bobwyman.pubsub.com/main/2005/10/erosion_of_powe.html">post</a> Bob Wyman talks of the “economy of information” this could enable. Good stuff:</p><blockquote><p>By providing the means for people and organizations to publish their structured data in the <a href="http://en.wikipedia.org/wiki/Surface_web">Visible Web</a> rather than the Gray Web or the walled-garden’s <a href="http://en.wikipedia.org/wiki/Hidden_web">Hidden Web</a>, we’ll be taking a great deal of “power” away from the walled-garden search sites and giving it back to the users and publishers.</p></blockquote><p>Bob’s scenario is only a small step away from what we’ve got now, but it’s still quite a significant step towards a Web of Data. With most current systems the integration aspects of this will be down to individual services, there isn’t a common data model. However pretty much any well-defined data can be interpreted in the <abbr title="Resource Description Framework">RDF</abbr> model (Structured Blogging, microformats, RSS/Atom, virtually any other XML, plus material currently hidden in SQL DBs…). As Tim Berners-Lee puts it:</p><blockquote><p>If HTML and the Web made all the online documents look like one huge book, RDF, schema, and inference languages will make all the data in the world look like one huge database.</p></blockquote><p>(<a href="http://www.w3.org/People/Berners-Lee/Weaving/Overview.html">Weaving the Web</a>, 1999)</p>
<<p><img src="http://dannyayers.com/2005/11/eric-sparql.jpg" alt="hedgehog and cat" /></p><p>Bit blurred I’m afraid, Eric trots off to find a dark place every time he’s out of his apartment. That’s his leg at the back. He seems to have long legs and heavy-duty feet, but is rather a nervous individual and it’s hard to have a close look - <em>ping!</em> He’s a ball.</p>
<<p>I’d forgotten that was coming ;-)</p><p>The <a href="http://www.croczilla.com/svg/samples/">Croczilla samples</a> (Tetris!) seem to work a treat. </p><p>The last bit of <a href="http://www.w3.org/Graphics/SVG/"><abbr title="Scalable Vector Graphics">SVG</abbr></a> I wrote was for <a href="http://www.wrox.com/WileyCDA/WroxTitle/productCd-0764570773.html">Beginning XML</a>, the main demo being a Tangram thing (here’s the <a href="http://dannyayers.com/archives/2004/08/06/new-book/">original code</a> plus some notes). Unfortunately it doesn’t work in FireFox (yet). For one thing I somehow managed to miss out the XLink namespace declaration (whoops!), but there are also a host of golden DOM incompatibilities. I found fixes for a lot of them <a href="http://svg-whiz.com/wiki/index.php?title=Cross-Platform_Authoring">here</a>, e.g.</p><p><code><del>event.getTarget();</del> <ins>event.target; </ins></code></p><p>Still haven’t got the dragging bit working (the half-fixed version is <a href="http://dannyayers.com/2005/11/tangram-cat.svg">here</a>). </p><p>What’s puzzling me a bit is that although it was a rush-job, I’m sure I checked all the code with both the Adobe plugin and Batik. Ah well.</p><p>Oh yeah, I guess it’s time to check out <a href="http://developer.mozilla.org/en/docs/Canvas_tutorial"><canvas></a> too.</p>
<<p>Gonna be offline for a bit, we’re having a new woodburning stove put in. The one we’ve got now is a stove in the US sense, a cooker, but we only used it for heating (did the living room and this office) and keeping soup warm. The new one will just be for heating, hopefully a lot more efficient (new chimney too) and less messy than the old one. It feels a bit strange, this is the kind of job we’d usually have done ourselves, but now way I’m getting on any loose-tiled roof until I’ve seen a local do it.</p>
<p><a href="http://dannyayers.com/archives/2005/12/01/grddl-job/">Danny Ayers</a> and <a href="http://dig.csail.mit.edu/breadcrumbs/node/29">Dan Connolly</a> both report the release of a new XML model for exchanging information between the the U.S. Department of Homeland Security (DHS), the U.S. Department of Justice (DOJ), and other government agencies. A final version of the NIEM is expected to be released in June 2006. </p><p>The current NIEM 0.1 is initially based on components and elements identified in the Global Justice XML Data Model (Global JXDM), Version 3.0.3, that is used by the criminal justice (Justice) domain. </p><p><a href="http://niem.gov/library.php">NIEM documentation</a>, <a href="http://niem.gov/documents/niem-0.1.zip">NIEM 0.1 package (zip)</a>, <a href="http://niem.gov/niem01.php">more information</a></p>
<p>Heute abend habe ich zum zweiten Mal in meinem Leben versucht in einem <a href="http://www.rossmann.de/servlets/Compose/SID=AAAj8babnnw1/help/10/4/c_index.html">Rossmann</a> einzukaufen, leider ohne Erfolg. Keine der Kassen wollte meine EC Karte akzeptieren - was ich schon sehr seltsam fand - und keine der Angestellten war in der Lage mir mit klaren Worten zu sagen warum es nicht geht… “Die Karte nehmen wir halt nicht, ist ne Mastercard”. Mal davon abgesehen das da Maestro draufsteht…</p><p>Ich mein… Hallo?? Ihr nehmt keine <a href="http://de.wikipedia.org/wiki/EC-Karte">EC Karte</a> der Deutschen Bank?</p><p>Also raus auss’m Rossmann, 4m weiter rein in’n <a href="http://www.dm-drogeriemarkt.de/">DM</a> und alles wird gut.</p>
<p>This week has been quiet busy with things in my day job… lots of communication and powerpointing. So, have not been able to hack a lot. :(</p>
<p>TiVo is rolling out new features that will allow users to discover new music, enjoy podcasts, purchase movie tickets, and view shared photos! Most of these features work with services from Yahoo!. If they work well, it will give TiVo new competitive advantage to against <strong>Apple</strong>, <strong>Microsoft</strong>, <strong>Sony </strong>and many others who want to enter the living room entertainment market.</p><p><a href="http://www.tivo.com/4.9.11.asp" title="TiVo">See the new features described on TiVo.com</a>. If you are a TiVo user, <a href="http://research.tivo.com/onlineservices/" title="TiVo sign up">go sign up here</a> for a faster update of your TiVo software.</p><p><img src="http://a423.g.akamai.net/7/423/1788/0c2425970332c1/www.tivo.com/i/4.0/4.9.11.YWeather.sm.jpg" /> <img src="http://a423.g.akamai.net/7/423/1788/bc1990e2ebd2a5/www.tivo.com/i/4.0/4.9.11.YPhotos_02.sm.jpg" /> <img src="http://a423.g.akamai.net/7/423/1788/9c4b23ca20019a/www.tivo.com/i/4.0/4.9.11.YTraffic.sm.jpg" /></p><p><strong>Update: <a href="http://apple.slashdot.org/article.pl?sid=05/12/02/2110234&from=rss" title=" Apple Enters Media Center Domain"> Apple Enters Media Center Domain</a></strong></p>
<<p><strong>7 reasons 2006 will be a big year for HTML</strong><br /><em>(with apologies to <a href="http://blogs.msdn.com/alexbarn/archive/2005/10/18/482515.aspx">Alex Barnett</a>)</em></p><ol><li><strong>HTML will ride the RSS slipstream.</strong> 2005 has been arguably the year of RSS. Since RSS is now ‘out there’ and a firmly established part of the internet, it won’t take nearly as long for HTML to reach a <a href="http://doc.weblogs.com/2004/03/10#publishing20">tipping point compared to RSS</a> in terms of time-to-critical-mass (RSS was invented in 1997, HTML in 1989). However, as we’ve learnt with RSS, in order for HTML to become a ‘mass use’ technology, <a href="http://blogs.msdn.com/alexbarn/archive/2005/07/19/440816.aspx">it will need to become ‘invisible’ to the user</a>.</li><li><strong><a href="http://www.oreillynet.com/rss/">Developers familiar with RSS</a> will ‘get’ HTML in a snap</strong>. The number of developers familiar with RSS is growing and will continue to grow, therefore the number of developers that know how to leverage HTML will grow with it - RSS and HTML are very related. Large <a href="http://www.internetnews.com/dev-news/article.php/3515406">software companies are playing with RSS</a>, so they will collectively ‘get’ HTML quicker than they ‘got’ RSS. Existing RSS-enabled and RSS-enhanced services and products can quickly enhance their offerings further by providing support for HTML.</li><li><strong><a href="http://blogs.msdn.com/alexbarn/archive/2004/09/03/225022.aspx">RSS is getting to the point of ubiquity</a>.</strong> Regardless of whether the users know they are using RSS or not, there appears to be a <a href="http://blogs.msdn.com/alexbarn/archive/2005/10/08/478598.aspx">critical mass of RSS users</a>. Getting to the situation where users to use HTML without them knowing they are using HTML doesn’t seems to be a huge stretch of the imagination.</li><li><strong>It looks like 2006 will be a healthy environment for start-up and internal project funding</strong>. <a href="http://blogs.msdn.com/alexbarn/archive/2005/10/05/477634.aspx">There has been a recent spate</a> of RSS-related acquisitions and VC-backed RSS-enabled start-ups emerging cropping up. With a willing investment market hungry for new opportunities it is a good time for HTML-related start-ups to get a sympathetic ear from those looking to invest in emerging syndication technology and media markets.</li><li><strong>There are two sides to the HTML coin: consumption and publishing</strong>. On the consumption side, there are <a href="http://allrss.com/rssreaders.html">more RSS readers you can shake a stick at</a> - most already provide some form of HTML support and extending that support to enable <a href="http://blogs.msdn.com/alexbarn/archive/2005/10/12/480364.aspx">some simple but high value scenarios</a> won’t take to much effort for the feed reader developers. New <a href="http://www.taskable.com/">HTML browsers are appearing too</a>. On the publishing side, blogs have been the primary vehicle driving RSS. Adding additional HTML publishing support by the blogware developers to enable these new and interesting scenarios shouldn’t seem like a huge stretch of the imagination.</li><li><strong>There are some <a href="http://www.cristianvidmar.com/compass/17/">really useful scenarios that could be enabled by HTML</a></strong>.<br />HTML is not in a situation where it is a solution looking for a problem to solve - some of scenarios outlined are genuinely useful and HTML makes these relatively easy to enable.</li><li><strong>HTML is cool, is simple and it works</strong> :-)</li></ol><p><em>Bonus points for anyone with good links to substitute.</em></p><p>See also: <a href="http://microformats.org/wiki/xoxo-brainstorming#Subscription_information">xoxo-brainstorming</a></p>
<<p>Elliotte Rusty Harold has <a href="http://www.cafeconleche.org/#news2005December3">posted</a> some slides/notes on syndication for a class he’s been teaching. Good stuff, and I’ve no doubt he covers a lot more than what’s in the slides, but here are one or two points I’d emphasize.</p><p>Quick one, I’d tweak one point on RDF from:</p><blockquote><p>Tries to do to AI what the Web Did to Hypertext</p></blockquote><p>to</p><blockquote><p>Tries to do to Data what the Web did to Hypertext</p></blockquote><p>Right near the start, he’s got the question/answer:</p><blockquote><ul><li>Is this the Web?</li><li>It is HTTP; and it is XML</li></ul></blockquote><p>Right on. Personally I think I’d rant on about HTTP for a while, noting that the content server-to-client part of syndication is essentially RESTful, which in part explains how RSS has got as far as it has. Somewhere here I’d insert (maybe ERH does when presenting) something about the polling aggregators do.</p><p>After some of the historical stuff he then leads onto Atom. It can be tricky to explain why it was needed when there is RSS to anyone that hasn’t spent time coding around the stuff. As I see it Atom has three key benefits over RSS 2.0: a clear, community-consensus spec; mandates the identifiers of the Web (URIs); the content model isn’t broken (apart from its general opaque messiness, escaped HTML in content is fundamentally flawed - check <a href="http://www.intertwingly.net/blog/2004/05/28/detente">silent data loss</a>).</p><p>After the formats there’s probably need to reference the XML-RPC based client-to-content server APIs used in many blogging tools. The RPC approach is a lot more tightly coupled than it needs to be, and using HTTP methods directly with XML documents is considerably less brittle (<em>which in part explains how RSS has got as far as it has</em>). When XML-RPC works, sure that’s fine, but when it doesn’t it’s <a href="http://geeks.opml.org/2005/12/03#a592">a devil to debug</a>. ERH says he didn’t have time to get to the Atom Publishing Protocol in his lecture - well, the protocol itself is taking rather more time than anyone expected… But as a way of unifying that side of publishing with RESTful HTTP+XML I think it will be significant in opening the door to a lot of innovation. Before we move to Web 2.0 we really should do Web 1.0 right.</p><p>Heh, I can’t fault his coverage of OPML:</p><blockquote><p><a href="http://www.opml.org/spec">No good spec</a> and that’s a problem</p></blockquote><p>Sometime very soon I need to prepare some slides on how syndication ties in with Semantic Web technologies - which, by virtue of being Web-based it does pretty nicely (FOAF is very significant, then there’s microformats, Structured Blogging, Google Base, Ning and Microsoft’s sync/sharing extension, GRDDL and Atom/OWL). But for the background stuff, hopefully ERH won’t mind to much if I pilfer a little of his material ;-)</p>
<<p>Melodramatic title, huh? Well it was either that or “Content Escaping Really, Really Sucks”. I just got a feed validator error from a <a href="http://dannyayers.com/archives/2005/12/01/users-in-charge/">post</a> of a couple days ago (and they will be another one here). The error goes like this: <em>“description should not contain script tag”</em>. Sounds reasonable. However I’m getting it because I have this: <script>, <em>escaped</em> in my <em>content</em>. I typed that into my WordPress posting form as &lt;script&gt; but that gets passed into a <description> element in my (RSS 1.0) feed, and that’s not allowed. I posted to the feed-validator list, curious how I should refer to <script>, and <a href="http://weblog.philringnalda.com/">Phil Ringnalda</a> responded: </p><blockquote><p>RSS 1.0’s <description> is in exactly the same state as both RSS 1.0 and 2.0’s <title> element: you quite simply cannot use the character “<” in them, because they are not HTML but virtually every consumer treats them as HTML. This very minute, there are consumers (at least one that I know of) which don’t use <content:encoded>, and which do treat <description> as HTML, and which don’t sanitize HTML, which are treating your entry as the start of a script element which never ends.<br />As you yourself say, “see silent data loss.”</p></blockquote><p>Yup, I <a href="http://dannyayers.com/archives/2005/12/04/presenting-syndication/">mentioned</a> that only today, though didn’t expect to hit the issue the way I usually use this setup, not today anyhow ;-) I certainly hadn’t considered the endless script… There’s irony or something there.</p>
<<p>Steve Gillmor, the guy who’s been plugging the notion of attention as a valuable resource, has dropped another term on the table. In <a href="http://blogs.zdnet.com/Gillmor/?p=189">his post</a> he talks of Gesture Banks, the Gesture Economy, GestureRank. It does sound very buzzwordy, and at first I thought it maybe just deserved a binary digital gesture, i.e. two fingers. As with attention, it’s not very obvious what he’s talking about.</p><p>So here’s my reading of what he’s on about. The gesture as the unit of attention. They are (heavily context-dependent) events. Gestures are just our intensional, directed interactions with the software. These communication acts contain in themselves valuable information. That information could be used to assist the person in their activities (e.g. with predictive search) or it could be used by marketeers, in a way it’s like a very wide broadening of AdSense. It reminds me a lot of <a href="http://frot.org/devlog/">Jo Walsh</a>’s setup to capture all her interactions with the computer. That should be good, usefully mineable data, for both the individual as well as for anyone the individual chooses to share it with. I’m pretty sure there’s nothing new in the notion of gestures along with computers (I bet there are thousands of papers in the user interface domain for starters), but there’s little to suggest it’s been systematically exploited in the Web domain.</p><p>Whether or not this is what Steve had in mind, I think there’s some practical use to this view. Take this point:</p><blockquote><p>Deriving GestureRank is therefore a function not only of who the gesturer is, but what is the nature (or type) of the gesture, and who or what group or domain it is directed toward.</p></blockquote><p>Like I said, Steve’s (probably intentionally) vague about what a gesture actually is. But if you take as an example clicking on a link, modelling the event usefully is non-trivial. Click action plus URI is simple in itself, then there’s the gesturer - to maximise the potential value you’d want that as informative as possible, FOAF’s probably a very good place to start. The group or domain could be modelled trivially as the URI being clicked, maybe that’s enough. But the nature (or type) of the gesture is where it gets a little tricky. Say the click is the user usubscribing from an RSS feed. That’s a kind of gesture of disinterest in the resource (and by extension things associated with the feed, like the person publishing the feed). In this case the type of the gesture will be statically bound to the URI - I wonder if that’s true in a more general case? </p><p>This gesture stuff will be appearing in quite a heavily interconnected set of relationships. By talking of GestureRank, Steve’s going into analysis of the accumulated stuff, which would probably bring in numerics. This is fairly high-dimensional stuff, more GestureVector than GestureRank. </p><p>At one point Steve’s talking of Google Desktop and suggests the gesture/attention as a user interface enhancement. Hmm, I have a question nearby. Because we’re good at learning repeated tasks, there’s some advantage in making user interfaces very predictable. This often outweighs the potential benefit of having a UI that changes dynamically to follow the user’s behaviour patterns. It may be faster to have a commonly-used menu item at the top of the menu, but the action of clicking on a particular item can’t be automated in a motory part of our brain if the menu order keeps changing. How much does this generalise? There’s obviously an advantage in using familiar tools. Does this also mean there’s an advantage in viewing the sea of information in a constant fashion?</p><p>Anyhow, I’ve a feeling that the notion of gestures in the context of attention could be very useful indeed. When I next have chance to look at <a href="http://esw.w3.org/topic/Attention_RDF">Attention RDF</a> I’ll have a play, based on these 5 minutes thinking aloud it looks like the Gesture Ontology will largely write itself. There are smart proxies and recorders (<a href="http://simile.mit.edu/piggy-bank/">PiggyBank</a>!) available, so the collection side should be straghtforward. The rest should be interesting…</p>
<p>Thanks to Damien for the cool questions <img src='http://www.johnbreslin.com/blog/wp-includes/images/smilies/icon_smile.gif' alt=':)' class='wp-smiley' /> </p><p><a href="http://www.mulley.net/2005/12/05/a-new-net-visioned-he-interview-with-john-breslin/">Damien Mulley?s Blog � Blog Archive � A new Net visioned he - Interview with John Breslin</a></p>
<p><a href="http://www.nytimes.com/2005/12/04/politics/04strategy.html" title="Bush's Speech on Iraq War Echoes Voice of an Analyst">NYT reports</a> that a recent speech given by President Bush was actually written by Dr. Peter D. Feaver, a Duke University political scientist. This information was <a href="http://ebiquity.umbc.edu/blogger/?p=419" title="On the importance of metadata">discovered</a> by reading the RDF metadata of a PDF document. </p><p>Not that I’m interested to learn about the real person that writes Bush’s speech, but I’m concerned about potential privacy issues that are associated with embedding metadata within digital documents. While Microsoft Word hasn’t yet started to embedded expressive metadata, such as RDF, in Word documents, its users already face <a href="http://wordprocessing.about.com/od/protectingyourprivacy/" title="Protecting Private Information in Microsoft Word Documents">many critical privacy issues</a>. I wonder what the world will be like once all our documents are tagged with semantic markups and ontologies?</p>
<<p>The recent troubles at the Wikipedia are described in this <a href="http://news.com.com/Growing%20pains%20for%20Wikipedia/2100-1025_3-5981119.html?tag=st.prev">CNET News.com</a> piece, along with Jimmy Wales response. From now on only named editors can create pages, so this should prevent dodgy pages appearing unnoticed (which was apparently what happened with a page saying a bloke called John Seigenthaler was possibly implicated in the JFK assassination). </p><p>This won’t do much to help ego-swamps like the <a href="http://en.wikipedia.org/wiki/Podcasting">Podcasting</a> page. But there it seems <a href="http://weblog.burningbird.net/2005/12/04/podcasting-history/">Shelley</a>’s put on her editing boots on that one.</p><p>One of the comments on the News.com piece caught my eye, <a href="http://lamammals.blogspot.com/">Len Bullard</a> says: <a href="http://news.com.com/5208-1025-0.html?forumID=1&threadID=12066&messageID=91326&start=-1">Use Semantic Web To Impute Importance for Notification</a>. As usual, the man has a point.</p>
<<p>Of interest to masher-uppers -<a href="http://www.aggregateknowledge.com/">AggregateKnowledge.com</a> is a site currently with two main offerings, one is :</p><blockquote><p><a href="http://www.wsfinder.com/">WSFinder</a> - A wiki-based directory of web services and APIs that people are using to create mash-ups.</p></blockquote><p>The other is :</p><blockquote><p><a href="http://www.aggregateknowledge.com:8080/wsrelater/Welcome.jsp">WSRelater</a> - A recommendation web service that you can add to your website in less than half a day.</p></blockquote><p>I don’t know about the “half a day” bit, but it looks a relatively straightforward thing, GET/POST+API key, with XML result sets. The recommendation part seems to come through path counting in a subgraph. They’ve got an interesting approach, it’s a kind of RDF lite, presumably backed by a RDBMS.</p><p>The nodes are strings, which are given one (or probably more) of 30+ types. There is only one relation: <em>related</em>. It’s directed.</p><p>Here’s one of their examples:</p><blockquote><p>(PERSON: John, GROUP: Software Forum) — “The PERSON John is related to the GROUP Software Forum”</p></blockquote><p>The similarity bit is done like this:</p><blockquote><p>…if a developer wants to find a GROUP that is similar to another GROUP by looking at all the people in common, the following three types would be used.</p><p> fromtype: GROUP<br /> totype: GROUP<br /> pivottype: PERSON</p></blockquote><p>I couldn’t actually try it out, not knowing what the database knows…if you see what I mean.</p><p>Heh, this is funny:</p><blockquote><p>The key to make this system have a network effect across site is there must be a uniform naming of items across sites.</p></blockquote><p>Indeed. That’s why URIs are Cool.</p><p>Oh, they’ve got a <a href="http://www.aggregateknowledge.com:8080/wsrelater/API.html#BestPractices">Best Practices</a> section, with some tips like:</p><blockquote><p>Always identify people with the following unique identifier</p><p> SHA1SUM(”mailto:xyz@abc.com”)</p></blockquote><p>Ok, that sounds more like it. But it does mean I’ve given already them some junk data.</p><p>~</p><p>Oddly enough I was thinking about something quite close to this the other day, around the blather about outliners. I do like the outliner UI paradigm, but there’s a snag in its utility in that in the usual implementation, you only have one relation type, i.e. “childOf” (unless you made the relations nodes, I did that in the tree view for <a href="http://ideagraph.net">IdeaGraph</a> - ok, but it loses some of the intuitiveness of the tree). I’m still after the ultimate Getting Things Done tool, and naturally I’d go for an RDF backend. But for it to be useful I believe there needs to be a variety of relations, one task may be a subtask of another, a child of a task may just be further description etc. It occurred to me that within a specific domain, you could usually infer the relation type from the types of the source and target of the relation. The AggregateKnowledge folks mention this kind of thing:</p><blockquote><p> (PERSON: John, PERSON: Paul) — “The PERSON John is related to the PERSON Paul”</p><p>Here “is related to” may mean “is a friend of” and be used in the context of a social network.</p></blockquote><p>In other words, if you know you’re in the context of a social network, then you can apply the rule:<br /><code><br />X type PERSON<br />Y type PERSON<br />X isRelatedTo Y<br />=><br />X friendOf Y<br /></code><br />For a project-oriented system like my GTD pipedream, the hierachical editor could have typed nodes, but no relation except childOf. The data could even be represented as HTML, with, e.g.<br /><code><br /><ol class="project">Feed Eric<br /><li class="task">get gloves</li><br /><li class="task">open his apartment</li><br />...<br /></ol><br /></code><br />Run that through XSLT to produce an RDF serialization, something like:<br /><code><br />_:a rdf:type :Project .<br />_:a rdfs:label "Feed Eric" .<br />_:b rdf:type :Task .<br />_:b :childOf _:a .<br />...<br /></code><br />Then run that through a rules engine along with:<br /><code><br />{<br />?x a :Project;<br />?y a :Task;<br />?y :childOf ?x .<br />} =><br />{ ?y :subTaskOf ?x }.<br /></code><br /><em>(Dunno if I got that right, I haven’t done that part of the <a href="http://www.w3.org/2000/10/swap/doc/Rules">tutorial</a> yet ;-)</em><br />Once that’s in the store, useful (SPARQL) queries and regular (RDFS/OWL) inference should be possible.</p><p>(spotter: <a href="http://blog.broadbandmechanics.com/">Marc</a>)</p>
<p>Because Wikipedia is editable by anyone, <a href="http://www.iht.com/articles/2005/12/05/business/wiki.php" title="Wikipedia weakness: Open-source, and open to abuse">some people complaint about its reliability</a>. I think Wikipedia is a great but not perfect tool — like many other things on the Internet. Allowing anyone to edit an encyclopedia is like free speech. Think about the gossip world. I think all Wikipedia users should be aware of this fact.</p>
<<p><a href="http://www.ietf.org/rfc/rfc4287">RFC 4287</a></p><p>(spotter: <a href="http://www.tbray.org/ongoing/When/200x/2005/12/05/RFC-4287">Tim</a>)</p>
<p><a href="http://www.boards.ie/">boards.ie</a> got its second mention on RTE Radio 1 last week; this time on Saturday’s “The Business” in a segment about broadband in Ireland.</p><p>You can hear the <a href="http://www.johnbreslin.com/blog/wp-content/20051203a.mp3">audio snippet</a> of this show.</p><p>Links:</p><ul><li><a href="http://www.rte.ie/radio1/thebusiness/">RTE Radio 1 - The Business</a></li><li><a href="http://www.boards.ie/vbulletin/showthread.php?t=2054857947">Boards.ie on the radio again - boards.ie/vbulletin</a></li></ul>
<p>The day has finally come, long anticipated Needle - SemaNtic BackbonE for European Digitial LibrariEs starts up.</p><p>Two European semantic digital library projects <a href="http://www.jeromedl.org/">JeromeDL</a> and <a href="http://www.brickscommunity.org/">BRICKS</a> gather together with joint initiative that aims to promote semantic web technologies in librarian network.</p><p>In the first step, we would like to gather <a href="http://wiki.needle.corrib.org/">community</a> that would aggregate ideas and scenarios for Needle, and disseminate the semantic digital libraries research back to librarian community.</p><p>Following the community building we will come with concrete actions for building European-wide open, semantic backbone for digital libraries.</p>
<p>Dies ist mehr ein “nice to have” drop und einfach mal ein Versuch, ob das auch mit einem blog funktioniert</p><p>Intranet Security Suite</p><p>Ein Konfigmanagement Enforcer, welcher alle Tools innerhalb eines Netzes gleichzeitig bedient, die auf Netzwerkebene moeglich sind. Nicht nur die Firewalls ALLER verwalteten Maschinen werden zentral administriert, sondern auch</p><p> * hosts.allow (soweit unterstuetzt)<br /> * hostfirewall (doppelt erwaehnt der vollstaendigkeit halber)<br /> * smb.conf (welches Netz,Host darf)<br /> * sshd_conf (kann auch host und user beschraenken)<br /> * Verwaltung ssh_keys<br /> * ldapserver (mit einem scheme den Zugang auf erlaubte Server beschraenken)<br /> * httpd.conf (RewriteRegeln fuer Zugriff von IP’s in vhosts includen (mehr als rudimentaer mittels hostgroups wird es vermutlich nicht gehen)) </p><p>Das Deployment koennte mit cfengine geschehen. Ein entsprechendes Frontend mit der gewuenschten Intelligenz muesste noch geschaffen werden. Einfachster Ansatz ein subversion checkin in eine verkettete Tabelle mit anschliesendem Deploy mittels cfrun.</p>
<<p>I needed a way of pointing from an arbitrary page on a website back to the homepage, the link communicating that relationship. The concept of homepage is pretty well time-tested, but there doesn’t actually appear to be a specified way of doing the markup. The particular use case was pointing from an archived blog post back to the front page. (<em>This will would then be GRDDLed into triples, and associate the foaf:Person behind the foaf:weblog home URI as the foaf:maker/dc:creator of the archived foaf:Document…</em>). </p><p>So I <a href="http://microformats.org/discuss/mail/microformats-discuss/2005-December/002293.html">pinged</a> the <a href="http://microformats.org/mailman/listinfo/microformats-discuss/">microformats</a> list, if anyone would know, they would. From the suggestions, the best way of doing it would seem to be to use :</p><p><code><br /><link rel="home" title="Home" href="http://url/of/home/page" /><br /></code></p><p>This isn’t on the HTML <a href="http://www.w3.org/TR/html4/types.html#type-links">link types</a> list, but since everyone’s going rel-crazy with <a href="http://www.w3.org/TR/REC-html40/struct/global.html#h-7.4.4.3">Meta data profiles</a> that isn’t such an issue. It will be rendered as a button in the Opera browser, and the Mozilla site navigation toolbar has supported it in the past (anyone happen to know if it can be turned on in FireFox?). If all this wasn’t justification enough, Mark Pilgrim also suggests the approach in <a href="http://diveintoaccessibility.org/day_9_providing_additional_navigation_aids.html">Dive into Accessibility</a>. Works for me.</p><p>~</p><p>Remotely related tip: if you keep getting “timed out” pages (like I did) in FireFox 1.5, enter :</p><p><code>about:config</code></p><p>in the address field, load that pseudo-page, then enter :</p><p><code>network.http</code></p><p>in the <em>Filter:</em> field. Double-clicking on the values allows you to change them. I’m not sure which specific value made the difference, but doubling the values for max connections, keep-alive etc seems to have done the trick.</p>
<p><a href="http://dannyayers.com/archives/2005/12/06/this-weeks-semantic-web-4/">Danny Ayers, Raw Blog : This Week’s Semantic Web </a>:<br />- Danny Ayers’s weblog. Sharing the information overload…<br />� RFC 4287 : The Atom Syndication Format<br />rel=?home? �<br />This Week?s Semantic Web</p>
<<p>Now I put my street cred on the line. I was born in 1964, and I think I had a happy early childhood. I can’t remember many events from then, but what I do remember are <em>atmospheres</em>. On my last Amazon book-buying spree a few weeks ago, on a whim I added a CD by the <a href="http://www.swinglesingers.com/">Swingle Singers</a>, a group that were around when I was a little ‘un. They’re described as being a jazz/classical fusion, they do <em>a cappella</em> versions of classical tunes: the (double) CD I got is Bach & Mozart. It’s not altogether traditional chorale, the sleeve notes refer to scat, the Jazz singing style, not (I don’t think) something toilet-related. Somehow the Bach is enhanced by their particular vocal treatment, it works. I’ve probably not heard their music for three decades, but the CD finally arrived today, and I just put it on. Back came one of those atmospheres. </p><p><img src="http://dannyayers.com/2005/12/champions.gif" alt="The Champions" title="his superpowers enabled him to pee over buildings..." /></p><p>The Swingle Singers were one of those groups (like John Dankworth and Cleo Laine), that would appear in the musical interlude of shows like Morecome and Wise. But the atmosphere for me is more that of some of the other television programmes of the late 60’s/early 70’s: The Champions, Randall and Hopkirk (Deceased), Jason King, The Avengers, The Saint, The Prisoner. Something very cool, futuristic and stylish. There’s also something of the Hammer Horror Camp Gothic, merged with the clinical science of the BBC Radiophonics Workshop, and naturally Dr. Who (<a href="http://www.jonpertwee.com/">Jon Pertwee</a>). </p><p><a href="http://www.bbc.co.uk/doctorwho/classic/episodeguide/seadevils/"><img src="http://dannyayers.com/2005/12/seadevil.jpg" alt="Sea Devil" title="quick! behind the settee!" /></a><br />Ok, this is no doubt a very personal interpretation, it is after all eight people going bee-po boo-pa biddy-peep… *</p><p>There is one more angle, that Caroline noticed - because there are a few fairly traditional renditions, overall the things sounds very Christmassy. Which is a bit nostalgic in itself. Ah, the smell of Lego.</p><p><em>* Brandenburg Concerto No.3</em></p>
<<p>The point made by <a href="http://www.ensight.org/archives/2005/12/05/web-20-companies-need-to-scale/">Jeremy Wright</a> <del>(currently 404ing)</del>, and supported by <a href="http://gigaom.com/2005/12/06/even-in-web-20-scale-size-matter/">Om Malik</a> on the need for Web 2.0 startups to consider scalability of their systems is valid, as far as it goes. But I believe the argument is, in part at least, misdirected and doesn’t goes far enough. Web 1.0 is the size that it is more because it has a scalable architecture than the performance of any local systems. The big issue isn’t that individual companies don’t build scalability into their own architectures, rather that they don’t tend to adequately exploit the scalability of the Web. Definitions of Web 2.0 vary, but I think two key aspects are <em>The Web as Platform</em> and <em>The Architecture of Participation</em>. For applications that exploit these paradigms to function, there must certainly be some forward-looking design locally, for example in setting up a distributed database on the company’s servers. But I believe more importantly there needs to be more work done at the edges, moving Web-common features to them, away from the application-specific aspect. </p><p>If the Web <em>is</em> the platform, then as much of the data that isn’t entirely application-specific must be exposed in such a way to allow it to be devolved to other parts of the Web. Use of existing Web-based data (i.e. other people’s exposed stores) shouldn’t be the exception, it should be the rule. This mean that for the system to be robust, there will be more work on caching, and less on the construction of disconnected data silos. Caching is one of the key features that has enabled Web 1.0 to scale. The current caching architecture is agnostic - it works now for human-oriented documents, there’s no reason it shouldn’t work for machine-oriented data. Sure, in general there is still work to do on data interchange and caching, but there’s already a lot in place and a lot more on its way (I personally believe the Semantic Web technologies offer a good route to commodity stores and distributed queries). </p><p>The mash-up is the prototypical Web 2.0 Service, but many current systems are much more enclosed, following the Web 1.0 walled garden pattern, jealously guarding parts of their systems which they perceive to be valuable, but which in the Web-wide marketplace have no distinguishing features. Things like identity management are best managed cooperatively, in general sharable data should be shared. For their to be real innovation in the Web space, startups and their developers should be free to concentrate on their Unique Selling Propositions, not reimplementing what is already widespread. The advantage in this approach is that they can be even more lightweight, there is significantly less work needed on infrastructure than in the Web 1.0 <em>Pet Store</em> mindset. </p><p>For an example of what I’m talking about here, consider the FeedMesh. A little consortium of blog search-oriented companies which share ping data arriving from frequently-updated sites. A part of that system is closely tied to a specific protocol and model, the data input, the ping endpoints. But the transfer of data between members of the consortium is not tied to the specific data application in question, it’s just a stream of XML. What members of the consortium do with that data is entirely up to them. In this case I believe that on change notification, they each go and collect the data from the remote sites, populating their local stores. But it isn’t difficult to imagine the feed data from those remote sites as being the currency streamed between FeedMesh members, in fact their has been discussion of it being done that way. Right now the local databases of these companies are probably implemented in widely divergent ways, tailored to the particular service that the companies offer. However, given that they are all receiving, forwarding, processing and storing the same kind of data, there’s no reason that they couldn’t each use the same commodity software for the non-service-specific parts of their operations. Ok, right now most of the FeedMesh companies are probably building on Apache, PHP and MySQL. But there’s significant commonality of their systems on a layer above. There is a shared data model, in this particular case that of syndicated feeds. The data exchanged is based on standards (e.g. Atom or RSS). At this point in time the FeedMesh itself is a relatively closed system, and in that sense not Web-friendly. But in the same way Intranets inform decisions for the Internet, so to can systems like these which have desirable features (another example might be Google Base). </p><p>So here’s how I imagine a possible future application architecture scenario. A startup’s software will consist of three parts commodity to one part unique. The commodity parts will be the bulk of their data storage, with standard interfaces. There are at least two architectures for this in the pipeline - for document/content/feed-related application, Atom Stores supporting the Atom Publishing Protocol should be eminently suitable. For more generic data, RDF stores with RDF/XML interchange and the SPARQL protocol and query language will be available (in fact an Atom Store could be built this way). The unique part of the startup’s system will be specific to their application, perhaps receiving data from other sources, user interface and processing this alongside the commodity data. They may well be generating new data, which would be fed back into the commodity store. For business reasons they’ll no doubt want to implement some access controls to ringfence some parts of the system. But for the most part, the commodity store and interfaces would be open to the rest of the Web, with data flowing to and from it freely. This would in effect be acting as a cache of Web data. The Architecture of Participation. Because this block of functionality would be something lots of people could use, the many-eyeballs mechanism of open source should ensure that robust scalability is built in. Managing expansion of the user base of the service will mean lower demands on system development (only 1/4 will need specialist development work), scaling up will primarily mean adding more commodity hardware.The fact that 3/4 of the system is commodity, and so could be available off-the-shelf will allow the company to innovate with the remaining 1/4. </p><p>In other words, the Web is the Platform, let’s use it. </p><p>See also: <a href="http://www.mnot.net/blog/2005/11/26/caching">Leveraging the Web: Caching</a>, <a href="http://www.tnl.net/blog/entry/Signs_of_a_Bubble">Signs of a Bubble</a> </p><p>PS. Richard MacManus <a href="http://blogs.zdnet.com/web2explorer/?p=74">says</a> he found this post interesting (thanks!) and put in a nutshell a point I was trying to make :</p><blockquote><p>…because commodity data is such an integral part of many Web 2.0 services, then caching in effect acts as a storage mechanism for data.</p></blockquote>
<<p>Ok, a rant.<br />Now I’m well into this Web 2.0 stuff, but there is one recurring idea that just plain irritates me. I believe it owes its current popularity to <a href="http://www.cluetrain.com/">the cluetrain manifesto</a>, a latter-day grimoire, that grants power to its readers not through direct interpretation but Gnostic contemplation. The manifesto is largely incomprehensible on face value, but in trying to make sense of it, Zen is achieved. And yes, like the New Age movement, cluetrain homogenises previously distinct religions (corporate public relations and Gonzo journalism), with similarly cheesy results. </p><p>But back to that irritant. Here’s a prime example of the way in which this irritant appears, in otherwise coherent <a href="http://www.imediaconnection.com/content/7486.asp">material</a> from John Battelle :</p><blockquote><p>“Online marketing is driven by conversation,” Battelle explained in his speech. In order to leverage the conversation, marketers and advertisers need to know the degree to which consumers are involved online under Web 2.0.</p></blockquote><p>If we take the word conversation literally, then this clearly isn’t true - even podcasts are usually monologues. But even if we take it metaphorically, it’s still not much use as an analogy. Conversations tend to be verbal, informal, free flowing, real-time, one-to-one at the water cooler or over the telephone or garden fence. Or small groups around the fireplace or propping up the bar. The corporation and the individual rarely communicate with any of these characteristics online. Blogging is not this mode of communication. Wikis are not this mode of communication. IRC probably comes closest, but as evinced by recent <a href="http://www.dltq.org/?p=780">backchannel-triggered brawling</a>, these probably aren’t the kind of conversations that benefit marketeers. Ok, certainly something similar to friendly conversation may be approached, and is probably desirable in the kind of information-driven economy many of the pundits are talking about. But if markets are conversations, then there’s the danger of making the inverse true. Your online interactions are valued in dollars, and interpersonal communication becomes little more than a facade, a portal even, through which the exchange of currency and snake oil takes place.</p><p>I guess my real grumble is that I think systems like the Internet are generally more valuable when profit isn’t the prime motivation. Don’t get me wrong, I don’t have a problem with the notion of making money off the Internet, far from it, indirectly that’s how I’ve been making a living for the past few years. But hearing the Web analysed in purely commercial terms (albeit couched in non-commercial language) not only seems fundamentally flawed - to paraphrase the saying, <em>knowing the price of everything and the value of nothing</em> - but also broken on the surface. Viewing every person that uses the network solely as a passive consumer to be <em>leveraged</em> into being a paying customer, is just so <strong><code>web.1.0.bubble.dot.com.<em>wrong</em></code></strong>.The path to sustainable financial profit through Web 2.0 isn’t that of search engine optimization via faux-conversation, but in building and joining together genuinely useful, interesting and entertaining systems. </p><p>I think Ambrose Bierce pretty well <a href="http://www.worldwideschool.org/library/books/lit/humor/TheDevilsDictionary/chap3.html">foresaw</a> in 1900 the current hype-driven abuse of a perfectly good word.</p><blockquote><p>CONVERSATION, n. A fair to the display of the minor mental commodities, each exhibitor being too intent upon the arrangement of his own wares to observe those of his neighbor.</p></blockquote>
<p>I think that doesn’t mean that <a href="http://www.rsasecurity.com/">RSA Inc.</a> has<br /><a href="http://flickr.com/photos/goern/71243277/"><img src="http://static.flickr.com/18/71243277_60dc3814f8_t_d.jpg" alt="RSA off" valign="top" /></a> been taken out of business… </p><p><em>PS</em>: This has been a <a href="http://en.wikipedia.org/wiki/Security_token">Token</a> given to me by my employer to log onto it’s <a href="http://en.wikipedia.org/wiki/VPN">VPN</a>.</p>
<p>I was showing off my iPod nano (won at the <a href="http://simile.mit.edu/">Simile</a> ISWC 2005 semantic bank competition) at the Cork geek bloggers’ dinner recently: just got <a href="http://idoom.hyarion.com/index.php">iDoom</a> working yesterday (thanks for the link to it Brendan!) and was watching videos on it today after installing the latest kernel from <a href="http://ipodlinux.org/">iPodLinux</a>. The video functionality is pretty nifty (but requires large AVI files because it is uncompressed); was playing the Superman Returns teaser trailer and the latest episode of the Simpsons (very watchable).</p><p>However, I have my first iPod nano screen scratch; I popped the nano out of its protective Jelly Belly box and was less than happy to see a small mark right in the middle of the screen. Hopefully the shop will take it back (I managed to retrieve the receipt, thanks Brian!).</p>
<p><a href="http://edition.cnn.com/2005/TECH/12/07/silicon.valley.garages.reut/index.html">Many Silicon Valley companies started their business in a garage</a> — HP, Google, Apple. I wonder if garage really has some kind of mythic power in helping technology business to be successful.</p>
<<p>I’m still trying to find/figure out nice ways of doing cross-site queries. One possibility that occurred to me that might be useful in fairly tightly-defined domains is something like this -</p><p>A <em>FOAFHelperServer</em> is defined as being as a Web server which supports the following:</p><p>1. (Mandatory) An endpoint URI which can answer queries of the form:<br /><code><br />CONSTRUCT<br />{ ?subject ?predicate ?object . }<br />WHERE<br />{ ?subject rdf:type foaf:Person . }<br /></code></p><p>2. (Optional) An endpoint URI which can answer queries of the form:<br /><code><br />SELECT ?endpoint WHERE<br />{ ?endpoint rdf:type x:FOAFHelperServer . }<br /></code></p><p>~</p><p>The fun bit. Say you wanted to answer the following <a href="http://www.w3.org/TR/rdf-sparql-query/">SPARQL</a> query:</p><p><code><br />SELECT ?name WHERE<br />{<br />?personA foaf:name ?name .<br />?personB relationship:collaboratesWith ?personA .<br />?personB foaf:name "Eric Vitiello" .<br />}<br /></code></p><p>The schema for the <a href="http://purl.org/vocab/relationship/">Relationship</a> vocab includes this:</p><p><code><br />relationship:collaboratesWith rdfs:label "Collaborates With" .<br /></code></p><p>Google supports wildcard queries, and it’s not hard to imagine a bit of string tweaking to change the above into a call on the <a href="http://www.google.com/apis/">Google Search API</a> looking like this:</p><p><code><br /><a href="http://www.google.com/search?q=*+Collaborates+With+Eric+Vitiello">* Collaborates With Eric Vitiello</a><br /></code></p><p>Now all you have to do is run a regexp against the result docs against the values of:</p><p><code><br />SELECT ?name WHERE<br />{<br />?person foaf:name ?name .<br />}<br /></code></p>
<p><a href="http://my.boards.ie/"><img src="http://www.johnbreslin.com/blog/wp-content/20051208a.jpg"/></a></p><p>Launched <a href="http://my.boards.ie/">my.boards.ie</a> this past week… I hope that it will act as a personalised area for communities and users of boards.ie. It runs vBDrupal and shares our existing member registration system. Also, I’ll soon be able to adapt my <a href="http://rdfs.org/sioc/drupal/">SIOC Drupal</a> plugin for it. There are two main thrusts to the site:</p><ul><li><b>Community Portals:</b> In this section, moderators can create articles about their communities on boards.ie.</li><li><b>User Blogs / Topics:</b> We have also enabled free blogging for registered users (with a reduced feature set compared to subscriber system).</li></ul>
<<p>Standards are extremely good things for the network, but as a data producer on the Web it can be difficult to conform to specifications, especially when most of the Web data consumers (browsers, aggregators) are extremely liberal. Problem notification comes late, usually when things are <em>really</em> broken. But there are good online validation services, in fact since they set up a <a href="http://validator.w3.org/feed/">enhanced clone</a> of the <a href="http://feedvalidator.org/">feed validator</a>, all the validators I personally need in the foreseeable future are available at the W3C. The main things are this HTML blog content, the same data as a feed and whatever RDF/XML I want to produce. </p><p>Right now (disregarding the fact that my <a href="http://dannyayers.com/feed/rdf/">feed</a> is enriched RSS 1.0) there’s not much RDF/XML from my “live” areas, and it’s relatively static: a <a href="http://dannyayers.com/misc/foaf/foaf.rdf">FOAF profile</a> and the site <a href="http://dannyayers.com/wp-sioc.php">SIOC</a> file. But before long I want to have something more dynamic online, along the lines of presence. This will probably build on the <a href="http://xmlns.com/foaf/0.1/">FOAF vocabulary</a>, adding bits from <a href="http://crschmidt.net/semweb/menow/">MeNow</a>, <a href="http://developers.technorati.com/wiki/attentionxml">Attention.xml</a> (expresssed as a <a href="http://micromodels.org/">micromodel</a>). I’m already tooled-up for some of this with things like <a href="http://pragmatron.org/trac/file/pragmatron/xslt/opml2skosroll.xsl">opml2skosroll.xsl</a> being available to snag my subscription data from <a href="http://www.bloglines.com/export?id=dannyayers">Bloglines export</a>. </p><p>Anyhow, the question is how to ensure all three kinds of data (HTML content, RSS/Atom feeds, RDF/XML presence) are valid. After dropping in a a little <a href="http://lists.w3.org/Archives/Public/www-validator/2005Nov/0115.html">tool request</a> to the relevant lists, it seems the W3C validators are <em>almost</em> at the point where it’s be straightforward to hang simple pass/fail services off of them, but right now would need a fair bit of work (the results are SOAPy for a start). I think for me the best way I think would be to either periodically run check (via cron) or trigger checks from actions (like posting to WordPress) and indicate the results on my blog itself as traffic light style buttons. The other main alternative would be to hitch these things into the WordPress editing UI, but that seems overly coupled to a single application (might make a nice plugin though…).</p><p>But right now any of this would mean a good few hour’s work (and a bigger priority is to go through the site config to try and prevent the thing from going down so often). So for now I’ve bookmarked the test results pages for my URIs at the <a href="http://feedvalidator.org/check.cgi?url=http%3A%2F%2Fdannyayers.com%2Ffeed%2Frdf%2F">feed validator</a> and <a href="http://validator.w3.org/check?uri=http%3A%2F%2Fdannyayers.com%2F">HTML validator</a> and put these into my “Daily” bookmark folder in FireFox. This is the folder I open in tabs in the morning, contains links to Gmail, Bloglines, Planet RDF etc. This will hopefully not be too intrusive on my workflow, yet maybe enough to raise the validity time proportion from probably 50% to something perhaps beyond 90%.</p>
<<blockquote><p>…I submit that the notion that “markets are conversations,” promotes bad hygiene.</p></blockquote><p><a href="http://homepage.mac.com/dave_rogers/GHD12-05.html#note_2495">Dave Rogers</a></p><p>(spotter: <a href="http://weblog.randomchaos.com/">Scott</a>)</p>
<h5>What’s the Geospatial Semantic Web?</h5><p>It’s part of the future Semantic Web that exploits geospatial technology to help people and computing machines to discover and share information. It’s also a web-centric information space that leverages Semantic Web technology for better dissemination of geospatial information.</p><h5>Why should we study the Geospatial Semantic Web?</h5><p>Breakthrough technologies often result from the cross-fertilization of different technologies. By studying the Geospatial Semantic Web, we can better understand the use of geospatial technology in Semantic Web applications and how Semantic Web technology can enhance the existing geospatial information systems.</p>
<<p>Earlier I followed a Tim Bray <a href="http://www.tbray.org/ongoing/When/200x/2005/12/08/Read-Write-Web">pointer</a> to Hal Stern, who <a href="http://blogs.sun.com/roller/page/stern?entry=web_2_0_in_three">suggests</a> that Web 2.0 can be summed up as “read-write web”. I disagreed with part of this in comments (<em>there is a lot it doesn’t capture…</em>) but I must admit he’s got something in this line:</p><blockquote><p>Once you start creating and consuming, you’re worried about metadata, relationships, rights, derivative uses, attribution, distribution and location of your bits.</p></blockquote><p>Stern’s got one of those nice Roller blogs that provides email subscription to comments. <a href="http://www.redmonk.com/sogrady">Stephen O’Grady</a> just dropped in a link to a <a href="http://www-128.ibm.com/developerworks/blogs/dw_blog_comments.jspa?blog=351&entry=81961">similar definition</a> from James Snell which I think I blogged before, but like The New Avengers bears repeating. It doesn’t capture <em>much</em> more (ok, arguably execution is a pretty important aspect of services…), but wins hands down for geeky elegance:</p><blockquote><p><code>chmod 777 web</code></p></blockquote>
<<p>Article in Nature - <a href="http://www.nature.com/news/2005/051205/full/051205-8.html">Social networks of rappers differ from all other human networks</a>.</p><p>Degrees of separation -</p><p>Film actors (i.e. Kevin Bacon): 2.5<br />Company directors: 3.6<br />High-energy physicists (i.e. <em>I forget his name</em>): 5.9<br />Rappers: 2.9</p>
<p>GML is a language that attempts to provide standard vocabularies for sharing and exchanging geospatial information. The definition of the language is very comprehensive. GML can be used to express extremely complex geospatial concepts.</p><p>However, GML falls short in being the right language for semantic representation. The root of the problem is that the expressiveness of GML is limited by the expressiveness of the XML language. For example, you can <a href="http://geoweb.blog.com/314091/">use GML to express a particular time instant</a>. However, it’s no easy to reference this defined time instance from a different context (e.g., in a different document).</p><p>On the other hand, if you use <a href="http://www.isi.edu/~pan/OWL-Time.html">RDF/OWL to describe a time instant</a>. It’s very easy to reference the defined instance from a different document. This makes easy for extending the description. For example, in one document you define a time instance (e.g.,<br /><div class="codesnip-container" >urn-x:t1</div><p>), and in the other you describe the same time instant as an instance of some event (e.g.<br /><div class="codesnip-container" >(urn-x:t1, rdf:type, evt:Event)</div><p>) and specify its a calendar/clock value (e.g.,<br /><div class="codesnip-container" >(urn-x:t1, tm:hasCalendarClock, “2005-12-02T12:09:93″)</div><p>).</p><p>I believe the ability to represent geospatial semantics is of great importance when building geospatial applications. Not only it will enable applications to share information, but also it will allow applications to better reuse information.</p>
<<p>US Air Force leaders have released a new mission statement:</p><blockquote><p>The mission of the United States Air Force is to deliver sovereign options for the defense of the United States of America and its global interests — to fly and fight in Air, Space, and Cyberspace.</p></blockquote><p><a href="http://www.af.mil/news/story.asp?id=123013440">Air Force Link</a></p><p>(spotter: <a href="http://linux.sys-con.com/read/161925.htm">Jeremy</a>)</p>
<<p>I just had my first FireFox 1.5 crash (no obvious cause, Win2k, one instance with maybe a dozen tabs, several running scripts). Not something that bothers me overmuch, but this in the Quality Feedback Agent is just plain annoying:</p><p><img src="http://dannyayers.com/2005/12/quality.gif" alt="contradictory signals" /></p><p>I’ve also still got a significant problem I had with the previous version of FF, when downloading a file the whole thing grinds to a halt for a minute or more… But wait, only *sometimes* - it just occurred to me to check if it was hammering the processor, so I fired up Task Manager and tried a download link. But this time it seemed to work ok, and there was no excessive processor load. Hmm, but MEM usage seems fixed at 260MB, is that ok? It says have only 256M on this laptop (that’s strange in itself, I thought I went to 512MB a couple of years ago), and have set a fixed disk cache - 512MB. <em>Not so relelvant if it’s now working…There is what should be adequate swap space on all the partitions, so I’m at a loss as to the cause. I installed a download manager in the hope that would help, but it doesn’t seem to have made any difference. For now I have to remember to copy the URI and use wget.</em></p>
<p>A <a href="http://www.washingtonpost.com/wp-dyn/content/article/2005/12/08/AR2005120802122.html" title="Spanish At School Translates to Suspension">kid got suspended</a> because he conversed with his friends in Spanish. This is completely ridiculous. Human languages are meant for communication. If people have the right to free speech, they should also have the right to choose whatever the language they wish to use to communicate. This kind of punishment reminds me of how native American Indian kids were forced not to speak their native languages in the American schools.</p>
<<p><em>As you can probably tell from the style (and lack of animal references) this is intended to go somewhere other than this blog. But I thought I might as well throw this first draft out here anyhow.</em></p><p>One of the biggest problems with current search technologies is that simple keyword-based queries often return vast amount of irrelevant information. The popular search engines index a huge information space, the Web, the indexing being based on text analysis. The index itself can be seen as a partitioning of the search space, with characteristic patterns found in the content determining the location of the divisions. In Google’s case the text analysis is augmented by exploiting page creator’s use of hyperlinks to infer relative importance of different pages (PageRank). With the index in place, queries are resolved by matching the terms they contain against the index. The indexing and matching takes place behind the scenes, all the user generally sees is a simple HTML query submission form, following by a list of the results. There are ways of making the search queries more explicit, for example using boolean combinations of keywords or restricting the search space to certain sites. But every new query is treating in the same fashion as the last and without any extra information than that provided in the query string. </p><p>Search personalization is the idea that individuals tend to be interested in specific topic areas, and that knowledge of their interests can be used to improve the relevance of the results the search engine returns. This knowledge provides a context through which search activities can be interpreted. There are various ways in which this knowledge could be obtained. An explicit profile of the user and/or their interests can be created as a static set of data, for example containing their geographic location. This would enable searches for real-world services to be restricted to the person’s local geographic area. However this approach demands effort on the part of the service user in creating the profile, and requires indexes within the service corresponding to the facets described in profiles.</p><p>A similar kind of information can be built up over a sequence of subqueries, each in turn narrowing the search space. Traditional directories, i.e. hierarchical taxonomic categorization, is one way through which the search space can be restricted. In the extreme case keyword matching isn’t needed, the classification allows narrowing of the information space through the user’s choice of sections and subsections of the topics of interest. This of course presupposes not only that the data is mapped by a classification scheme, but that the scheme used is appropriate for the user and their specific queries. Experience shows this to be a difficult problem, especially when an information space as large and as dynamic as the Web is to be considered. The choice of classification scheme is fairly arbitrary, and manual classification is prohibitively time-consuming. The editors of the Open Directory Project [<a href="#1">1</a>] are overwhelmed by the number of pages awaiting classification, and although Yahoo! had a relatively successful directory in the earlier days of the Web, the acceleration in the growth of the Web soon pushed them towards free-text search. </p><p>A more ambitious approach to search personalization is to monitor and analyse the search client’s behaviour over time and from this infer information about the nature of the information the person is looking for. This approach is very attractive from the end user’s point of view, in that no preparation is required, and no drilling down through categories for every individual search activity. But this will be at least at the cost of increased complexity in server-side systems, being in essence an AI-style machine learning task. It has been argued that this approach to search personalization is a dead end [<a href="#2">2</a>], but there is a lot to suggest that when considered in concert with other techniques, there are side avenues that may potentially prove very rewarding. </p><p>Perhaps part of the reason search personalization is described as a dead end is because it appears too hard to do usefully. To date, search and cataloguing systems have tended to take relatively narrow approaches, no doubt because integration of data across multiple indexing/classification schemes and user interface paradigms has traditionally been difficult. But techniques have been evolving in recents years, notably the Semantic Web technologies, and these can relieve many of the harder issues, especially with data that is sourced from the Web. There is no magic bullet, but these systems can help integration between text-oriented indexing and structured taxonomies. Perhaps more significant is the way these technologies enable easy integration of traditional search/catalog data with other sources of potentially very relevant information, such as “folksonomic” keyword tags, personal profiles and social relationships. With the addition of syndication technologies and IM, there is the prospect of continuous, real-time evolution of Web-based information systems. </p><p>Techniques and tools for capturing certain aspects of user behaviour are becoming available. Specifically for search Google’s own Personalized Search [<a href="#3">3</a>] is an end-user tool which records the individuals search activity, allowing subsequent searches to be filtered and/or ranked with reference to prior activity. (In Google Labs there also appears to be a prototype profile-based system [<a href="#4">4</a>] where the user may select topics of interest, presumably to assist in the filtering/ranking of results). However Google’s tools do not appear to be openly available to developers.</p><p>Outside of the search-specific arena, the term “Attention” is being used to described the focus of a particular person at a particular point in time. This has been recognised as potentially valuable information, to the extent that there’s an organisation, AttentionTrust, [<a href="#5">5</a>] devoted to encouraging practices which give the individual control over their own attention data. Examples of tracking techniques associated with this view of behaviour are include Attention.xml [<a href="#6">6</a>] a format specification which encapsulates a domain model in which user interaction with RSS/Atom feeds can be captured, and AttentionTrust’s Attention Recorder tool [<a href="#7">7</a>] (a FireFox plugin) which records user interactions with their Web browser. The data from the Attention Recorder can subsequently be uploaded to online services, for example Root Vaults [<a href="#8">8</a>] offers some basic tools for examining the data.</p><p>But even assuming plenty of data surrounding the user’s behaviour is collected, the problems of personalized search based on machine learning of historical activity are still not trivial. Specific problems identified (paraphrased from [<a href="#2">2</a>]) include:</p><ol><li><strong>People are not static</strong><br />The issue here is that people’s interests change over time, the topics in which they are interested is by no means a fixed list. Their interests while working will relate to their profession, in their leisure time they may have a multitude of fairly unrelated hobbies or interests. However, even if the set of topics is changing over time, a large proportion of an individuals search queries will be clustered around specific topics. This problem does not rule out the ability to recognise increased (or decreased) relevance of results found in particular domains. Things like homonym disambiguation are still possible. For example, if a person has previously used the keyword “rowing” in queries and visited sites relating to boats, then chances are they will be thinking of the same context when they use the word in future, rather than domestic disputes. What this problem does mean is that any system designed to track and make inferences from attention will need to take the dynamic nature of people into consideration. One important point here is that pages about topics of interest appear in clusters in the information space, and one area of interest may be totally distinct from another. Attempts to “average” or linearise results over time without taking this into consideration are not likely to produce useful results.</li><li><strong>The surfing data used for personalizing search is weak</strong><br />The point here is that Web surfing behaviour often involves little commitment, a single click can take a user to a page, another click can take them away again. The contrast is made with buying a product at Amazon. But this doesn’t mean that this kind of data is useless, only that what’s collected will contain noise. There are well-establish statistical and engineering algorithms for extracting useful information from noisy data. What’s more, with enhancements to the tools they use, the user can easily make the data collected while surfing considerably stronger. For example, the <a href="http://simile.mit.edu/piggy-bank/">PiggyBank</a> Firefox plugin can assist the user in extracting explicit data from Web pages, and maintaining it in a personalised store.</li><li><strong>Users interact based on limited information</strong><br />When a search engine user clicks on a result link, they are doing so not because of the material on the remote page, but based on whatever information the search engine has displayed. This is likely to be another source of noise, and the user behaviour may well be skewed by the way in which the results are presented. But again, number crunching can be used to circumvent noise and skew in data.</li><li><strong>Computers tend to be shared</strong><br />A home computer, or even one in some office environments might be used by a variety of people. Clearly the mixing of attention/personalization data from different people would seriously reduce its utility, in fact it breaks the whole point of personalization. However it is already often in the interests of the individual user to identify themselves to the system to use for example a personalized desktop, in fact if authentication is required then self-identification may be obligatory. If one assumes personalized search has significant value, then it will be in the users interests to enable it, by somehow signing themselves in and out.</li><li><strong>Queries tend to be short</strong><br />Short queries may provide little contextual information in themselves, but this is probably the strongest argument for collecting <em>more</em> user behaviour data, so that such queries can be contextualized.</li></ol><p>Aside from these issues, there are naturally the purely technical considerations. If we are talking in terms of accumulating user data from a variety of sources (their browser and aggregator use, for example) and using that alongside personal and topic-oriented profiles, in a system that needs to carry multiple index and taxonomic schemes, and from all this producing useful results, we are talking of what would traditionally be a data integration nightmare. As suggested earlier, it is likely that Semantic Web technologies will be very useful here, as they are designed to enable integration and inference over the kind of data the Web has to offer. </p><p>There are many significant issues surrounding personalized search not touched upon here such as performance and privacy, and many of these issues are not yet resolved. But the issues raised above don’t convincingly negate the value of personalization, if anything they suggest why it would be valuable. The Vivissimo document from which this problem list was derived concludes:</p><blockquote><p>The best personalization is done by persons themselves.</p></blockquote><p>Absolutely. But there’s no reason the computer shouldn’t be able to assist them.</p><p>[<a name="1">1</a>] <a href="http://dmoz.org">Open Directory Project</a><br />[<a name="2">2</a>] <a href="http://vivisimo.com/docs/personalization.pdf">Why Search Personalization is a Dead End</a><br />[<a name="3">3</a>] <a href="http://www.google.com/psearch">Google Personalized Search</a><br />[<a name="4">4</a>] <a href="http://labs.google.com/personalized/profile.html">Google Lab Profiles</a><br />[<a name="5">5</a>] <a href="http://www.attentiontrust.org/">AttentionTrust</a><br />[<a name="6">6</a>] <a href="http://developers.technorati.com/wiki/attentionxml">Attention.xml</a><br />[<a name="7">7</a>] <a href="http://www.attentiontrust.org/services">Attention Recorder</a><br />[<a name="8">8</a>] <a href="http://root.net/vaults">Root Vaults</a></p><p>See also : <a href="http://en.wikipedia.org/wiki/Attention_economics">Wikipedia : Attention economics</a></p><p><a href="http://www.ratcliffeblog.com/archives/2005/12/community_perso.html">Mitch Ratcliffe</a> :</p><blockquote><p>Meanings are <strong><em>refined</em></strong> by the community rather than <em>defined</em> by the community. Personalized search results are intricately linked to the evolution of the online community, so the individual’s attention to particular communities and information is essential to parsing available information to their interests.</p></blockquote>
<<p><em>[A quiet week…]</em></p><h4>Events</h4><ul><li>Streaming video from <a href="http://www.swap2005.org/">SWAP2005</a>, 15th and 16th December</li><li><a href="http://www.xtech-conference.org/2006/call.asp">XTECH 2006</a> <em>“Building Web 2.0″</em>, 16-19 May 2006, Amsterdam, Netherlands</li></ul><h4>Docs etc</h4><ul><li><a href="http://my.opera.com/chaals/blog/">Chaals has a blog</a>!</li><li><a href="http://lists.w3.org/Archives/Public/semantic-web/2005Dec/0052.html">Call for contributors</a> to the “Semantic Web FactBook 2005″ & “SW @ your country reports”, from the editor of AIS SIGSEMIS Bulletin (<a href="http://www.sigsemis.org/newsletter/december2005/vol2-issue34.pdf/view">latest issue</a>)</li><li>Leigh Dodds has been <a href="http://www.ldodds.com/blog/archives/000261.html">looking at XMP</a>, where <a href="http://www.adobe.com/products/xmp/main.html">XMP</a> is Adobe’s embeddable subset of RDF (as found in the output of all their tools)</li><li><a href="http://lists.oasis-open.org/archives/office/200512/msg00009.html">Discussion</a> continues over the suitability of XMP for <a href="http://www.oasis-open.org/committees/tc_home.php?wg_abbrev=office">Open Document Format</a> metadata</li><li><a href="http://www.idea-group.com/jwsr">Journal of Web Services Research</a> (JWSR), with submission guidelines</li><li>CfP : <a href="http://www.globalintegrationsummit.com">Global Integration Summit</a> (GIS’06) <em>“Integration For Everyone”</em>, Boston US 22-24 May 2006</li><li>In the blogosphere #1, Chris Eidhof <a href="http://www.my-website.nl/weblog/2005/12/11/semantic-html-rdf-and-the-semantic-web/">discusses Semantic HTML and RDF</a></li><li>In the blogosphere #2, Nassib Nassar on <a href="http://www.etymon.com/wordpress/?p=7">Federated Databases in Science</a></li><li><a href="http://www.markboulton.co.uk/journal/comments/semantic_typography_bridging_the_xhtml_gap/">Semantic Typography: Bridging the XHTML gap</a></li><li>Alongside articles on <a href="http://www.xml.com/pub/a/2005/12/07/catching-up-with-the-atom-publishing-protocol.html">Atom Protocol</a> and <a href="http://www.xml.com/pub/a/2005/12/07/handling-atom-text-and-content-constructs.html">Atom text handling</a> there’s the <a href="http://www.xml.com/pub/a/2005/12/08/2006-xmlcom-reader-survey.html">xml.com reader survey</a></li></ul><h4>Software and stuff</h4><ul><li><a href="http://www.flickr.com/groups/35237098612@N01/discuss/86234/">Flickr2FOAF</a> : <a href="http://f14web.com.ar/inkel/2005/08/25/flickr2foaf.en.html">Python scripts</a> from <a href="http://f14web.com.ar/inkel/">inkel</a></li><li><a href="http://www.dbai.tuwien.ac.at/proj/semnum/phoaf.html">PHOAF #1</a> - The SemNum Prototype, putting ENUMs on the Semantic Web</li><li><a href="https://gna.org/projects/phoaf">PHOAF #2</a> - PHOAF: a PHP library to get informations from FOAF files</li><li><a href="http://www.w3.org/2004/02/skos/core/owl-dl/skos-core-owl-dl.owl">SKOS Core in OWL DL</a> - <em>“unofficial”</em></li><li><a href="http://web2.wsj2.com/the_best_web_20_software_of_2005.htm">The Best Web 2.0 Software of 2005</a></li></ul><p>Yep, a quiet week, perhaps time to take <a href="http://www.achewood.com/comic.php?date=08182003">a “me” day</a>… </p><p><em>Sources include <a href="http://planetrdf.com">Planet RDF</a>, <a href="http://journal.dajobe.org/journal/2003/07/semblogs/">Semantic Weblogs</a>, <a href="http://swig.xmlhack.com/">Semantic Web Interest Group IRC Scratchpad</a>, <a href="http://www.w3.org/2001/sw">W3C Semantic Web Activity</a>, various emails etc - thanks!</em></p><p><em><a href="http://dannyayers.com/archives/category/virtual-world/semantic-web/sw-weekly/">other weeks</a></em></p>
<p><a href="http://dannyayers.com/archives/2005/12/11/this-weeks-semantic-web-5/">Danny Ayers, Raw Blog: This Week’s Semantic Web </a>:<br />- Danny Ayers’s weblog. Sharing the information overload…<br />This Week?s Semantic Web</p>
<<p>41 tasty papers - <a href="http://sunsite.informatik.rwth-aachen.de/Publications/CEUR-WS//Vol-166/">CEUR-WS.org/Vol-166 - SWAP, Semantic Web Applications and Perspectives, 2nd Italian Semantic Web Workshop</a></p>
<<p>I’m off to Trento tomorrow, don’t know how online I’ll be for the rest of the week, not taking a laptop (battery’s dead, not insured). Problem is this Apache2 server seems to keep locking up on me. I haven’t time to delve deep, but as a workaround could do with maybe a twice-daily restart. For some reason <code>apache2 -k restart</code> doesn’t seem to do the trick, I have to <code>killall apache2</code> first. I’ll try and figure out the script I need to attach to cron later (I think it’ll need to check all the apache2 processes have gone before restarting), but I’m a bit pressed for time and not in the least 133t so if anyone has anything suitable on hand I’ll be very grateful.</p>
<p><img align="left" alt="Nuclear Facility Iran" title="Nuclear Facility Iran" src="http://www.geospatialsemanticweb.com/wp-content/images/arak.nuke.plant.jpg" />One picture is worth a thousand words. But how do you capture that thousand words in a machine processable format so that the knowledge can be reasoned over, shared, searched, and archived? In the intelligence community, analysts are faced with this problem everyday.</p><p>At present, analysts rely on text reports to share intelligence information. Often this makes the sharing of intelligence information very difficult. Let’s take imagery analysis as an example. An analyst typically studies satellie images and analyzes the geographical features that are depicted in these images. Based on his/her knowledge, the analyst attempts to extract useful intelligence information from the analysis. For example, seeing the development of new military arm forces in a previously abandoned <span style="font-weight: bold"></span><span class="def-word"></span><span style="font-weight: bold">nuclear facility in North Korean, the analyst concludes that the country is attempting to reopen the nuclear facility.</span></p><p>Now, let’s assume the analyst writes down his/her conclusion in the report, and passes on this report to some other analysts. Based on the report, why should these analysts believe the author? Why should they believe that there is new military arm forces in the target region? What exactly are those geographical features or changes depicted in the pictures that made the analyst to draw his/her conclusion?</p><p>Even if we assume the author did write down a comprehensive description of his analysis, how easy would it be for this report and its content to be searched by different analysts in a later time?</p><p>It’s clear that plain text is not the best format for building up machine processable knowledge. To better facilitate this kind of intelligence analysis, there is a need for image annotation tools.</p><p>Some interesting image annotation tools and resources:</p><ul><li><a title="PhotoStuff" href="http://www.mindswap.org/2003/PhotoStuff/">PhotoStuff</a> — an application that allows the user to annotate different parts of an image with RDF descriptions.</li><li><a title="iPhotoRDF" href="http://www.holygoat.co.uk/applications/iphoto-rdf/iphoto-rdf">iPhotoRDF</a> — an Mac iPhoto plugin that allows the user add RDF annotations</li><li><a title="DOM Image Annotation Guide" href="http://rds.yahoo.com/_ylt=AsZlV_jh14p78SdmBA0x5NtXNyoA;_ylu=X3oDMTE2MW02Ym9kBGNvbG8DdwRsA1dTMQRwb3MDMQRzZWMDc3IEdnRpZANTTkdZMl8x/SIG=12cp3b8fc/EXP=1134514045/**http%3a//www.kryogenix.org/code/browser/annimg/annimg.html">DOM Image Annotation Guide</a> — a guide that reveals how Flickr builds its photo annotation capability using DHTML and Flash</li><li><a href="http://esw.w3.org/mt/esw/archives/cat_image_annotation.html">ESW Image Annotation Archives</a> — a list of image annotation tools that build on Semantic Web technology</li></ul>
<<p><a href="http://norman.walsh.name/2005/12/12/vcard">Norm Walsh</a> has provided an ontology, a welcome update to the pretty obsolete <a href="http://www.w3.org/TR/vcard-rdf">2001 vocab</a>. He also provides the <a href="http://norman.walsh.name/2005/12/12/examples/hcard2rdf.xsl">XSLT</a> suitable for use with <a href="http://www.w3.org/2004/01/rdxh/spec">GRDDL</a> that extracts vCard RDF from HTML marked up with the <a href="http://microformats.org/wiki/hcard">hCard</a> microformat. </p><p>He mentions the ontology etc. going to live with the W3C, which makes sense. Hopefully it’ll get a stable namespace URI early on.</p><p>This had been on my own to-do list for quite a while, at least the address bits. But I was hoping someone might get there first…thanks Norm! </p><p>Now where is that <a href="http://tantek.com/log/2005/09.html#d18t1657">Avon Lady</a> data…</p>
<p>Last week I constructed a <a href="/datenbrei/kontakt/">contact page</a> and enriched it with the <a href="http://microformats.org/wiki/hcard">hCard microformat</a>. Using the wonderful <a href="http://suda.co.uk/projects/X2V/">web service X2V</a> of Brian Suda anyone is able to retrieve a vCard file from my contact page. Now, yesterday the next wunderful Norman Walsh jumped in and reworked the vCard Vocabular and documented his work on <a href="http://norman.walsh.name/2005/12/12/vcard">Extracting vCards from hCard markup</a>. Norman’s work resulted in a <a href="http://nwalsh.com/rdf/vCard">RDF Vocabulary</a> and a <a href="http://norman.walsh.name/2005/12/12/examples/hcard2rdf.xsl">GRDDL transformation</a> (I have local copies of this in <a href="http://b4mad.net/2005/12/12/">here</a>). </p><p>Having all this in place there are several ways to represent my contact data: as <a href="/datenbrei/kontakt/">XHTML with hCard</a> microformat, as <a href="http://suda.co.uk/projects/X2V/get-vcal.php?uri=http://b4mad.net/datenbrei/kontakt/">vCard</a> and <a href="http://www.w3.org/2000/06/webdata/xslt?xslfile=http%3A%2F%2Fb4mad.net%2F2005%2F12%2F12%2Fhcard2rdf.xsl&xmlfile=http%3A%2F%2Fb4mad.net%2Fdatenbrei%2Fkontakt%2F&transform=Submit">vCard RDF</a>. Thanks to all for the work!</p>
<p>In the past, when the term “geospatial” is mentioned, people immediately think digitial maps. Today most people think Google Maps and Google Earth when the same term is mentioned. To me, seeing mapping technology as the sole component of geospatial technology is a nearsighted vision. <span style="font-style: italic" /><em><span class="def-word" /><span style="font-weight: bold" /></em></p><p>Geospatial technology is more than just pretty maps. <a title="IDC Reveals Radical Changes in Spatial Information Management That Will Impact Most IT Companies" href="http://www.crm2day.com/news/crm/116683.php">A recent IDC study</a> shows that the spatial information management industry is undergoing radical technology changes, which is likely to impact many IT ecosystems.</p><blockquote><p>Fundamental shifts in the spatial information management industry<br />include basic changes in the nature of geospatial work, and transitions<br />in the broad IT environment toward easier integration and support for<br />business processes.</p><p><u>The study finds that geospatial data, and not the map, has<br />become the raw resource for creating location-specific information.</u><br />Therefore, efforts to convert paper maps to digital data have been<br />replaced as geospatial data is used to generate new maps, decisions,<br />and automated processes.</p></blockquote><p>Let me take things one step further. I think a wide adoption of geospatial technology in IT is only the begnning. Some of the most exiciting applications in the future will be the ones that exploit <em>geospatial semantics</em>, not just geospatial data.</p>
<p><a href="http://sciencecommons.org/data/neurocommons" title="The NeuroCommons">NeuroCommons</a> is a Creative Commons project that is aimed to build the largest online knowledge base of neurological research. The project attempts to address the problem of hidden knowledge in unstructured scientific publications.<br /></p><blockquote><p><em>The backbone of the NeuroCommons is the scientific canon, or set of facts published in the neurological research. These connections - “this gene causes Huntington’s Disease” - represent the knowledge collected, reviewed, and published by scientists over decades of research. At the moment the vast majority of these facts are trapped in document formats that are readable only by individuals - PDF, Word, HTML - and in many cases, usage is constrained by copyright…</em></p></blockquote><p>I like the idea of building machine processable knowledge bases using Semantic Web technology. But I doubt that it’s feasible to use RDF to represent complex knowledge — such as the facts in neurological research. I think a more feasible approach, given today’s technology, is to annotate existing documents with ontological descriptions, and build intelligent applications on top of these metadata. </p>
<p>As the size of the Web gets bigger and bigger, search engines such as Yahoo! and Google may be too general for building applications that focus on some particular domain of information. To solve this problem, <a href="http://www.alexa.com/" title="Alexa">Alexa</a> provides a web search platform that allows people to define their own search engine. </p><p>Although you have to pay for the service, but it definitely looks promising. Alexa crawl works over 100 Terabytes of Web content spanning 4 billion pages and 8 million sites, and support a wide variety of types of content from the Web (jpgs, gifs, mp3s, movies. text/html, and even metadata). </p><p><a href="http://websearch.alexa.com/static.html?show=webtour/start" title="Alexa Quick Tour">How does Alexa work?</a></p>
<p>Three Star Trek stories caught my eye in the past day:</p><ul><li><a href="http://www.hollywood.com/news/detail/id/3467643">William Shatner Plans ‘Star Trek’ Prequel</a></li><li><a href="http://tcal.net/archives/2005/12/16/three-star-trek-captains-in-one-film/">Three Star Trek Captains in One Film</a></li><li><a href="http://www.wired.com/wired/archive/13.12/startrek.html">To Boldly Go Where No Fan Has Gone Before</a></li></ul><p>Of these, the last is still the most interesting, especially if you’ve watched any of the fan-produced <a href="http://www.newvoyages.com/">New Voyages</a> episodes. Walter Koenig (Chekov from the original series) is playing an older version of himself in the next episode!</p>
<p>As <a href="http://www.stefandecker.org/blog/">Stefan</a> recently blogged, we’ll be giving a tutorial at <a href="http://www.www2006.org/">WWW2006</a> as described below.</p><p><a href="http://www2006.org/tutorials/#T13">WWW2006 - Tutorials</a></p><blockquote><p><b>Semantic Web 2.0: Creating Social Semantic Information Spaces</b></p><p><i>John Breslin, Stefan Decker</i></p><p>This tutorial will give an overview of current proposals in the Semantic Web area for adding semantics to emerging and established communications media such as blogging and wikis. We will also cover the usage of Semantic Web technologies for community portals. We will discuss current standardisation activities as well as research prototypes. Additional topics to be covered include semantic search based on metadata and large scale data integration as well as semantics in digital libraries. Finally, we will discuss and present current approaches to realise the ideas of Vannevar Bush and Doug Engelbart on distributed collaboration infrastructures, which we term Social Semantic Information Spaces.</p></blockquote>
<p>Prof. Max Egenhofer has written a short paper, “<a title="Toward the Semantic Geospatial Web" href="http://www.spatial.maine.edu/~max/RC50.html">Toward the Semantic Geospatial Web</a>“, that discusses some key issues in building a new Web that can exploit geospatial semantics. He believes that in order for Semantic Geospatial Web (or geospatial semantic web as I call it) to take off, it will require <u>the development of standard geospatial ontologies for representing data and standard query languages for accessing data</u>.</p><p>I believe standard ontologies and query languages only solve part of the problem. In real world geospatial applications, building standard vocabularies and queries langugaes are the easy part of the tasks. The hard part of the problem is <em>how to integrate mass amount of geospatial data that already exists</em>.</p><ul><li>How can we integrate existing geospatial data without needing to create new databases that basically replicate the existing ones?</li><li>How can we query the semantic knowledge that is fused from heterogenous data sources without needing to know the specific representations of these data sources ?</li><li>How can we track the pedigree and provenence of geospatial data in a Web-based information space in which anyone can say anything about everything?</li><li>How can we faciliate the sharing of different types of geospatial data (images, videos, maps etc) in a Web-based environemnt?</li></ul>
<p>The Consumer Price Index measures the cost of goods that typical consumers purchase to live their everyday life. It’s a tool for measuring inflation. The <a href="http://www.pncchristmaspriceindex.com/index.htm" title="PNC Chrismas Price Index">Christmas Price Index</a> look at the increasing cost of goods and services bought by the True Love in the holiday classic, “The Twelve Days of Christmas.”</p><p>According to the index, the cost of seven swans a swimming has increased 20% over 2004. Similarly, six geese a laying is up 42.9%, and the partridge itself is up 12.9%. <a href="http://news.morningstar.com/doc/article/0,1,150214,00.html">Read more</a>.</p>
<p><!-- added by [GNU:] on 2005-07-11 --><br /><!-- inspired by a post on planetweb20.com --></p><blockquote><p><em class="strong">This is a Web 2.0 Website!</em> “Web 2.0 is an attitude not a technology. It’s about enabling and encouraging participation through open applications and services. By open I [Ian Davis] mean technically open with appropriate APIs but also, more importantly, socially open, with rights granted to use the content in new and exciting contexts.” — Ian Davis, 2005</p></blockquote><p>But from my point of view there is no such thing like Web 2.0, we got all the technology, all the tools, what I encorage is the attitude described above.</p>
<<p>Got back from Trento last night, great conference (grazie!). More later, for now got a *lot* of catching up to do…</p>
<<p>Comments. Oops. At least 6 of the comments in the moderation queue on this blog should’ve gone straight through. Apologies. Probably calls for a separate post to address them - when I’ve caught up a bit more.</p><p>This comment I blocked, but still like it:</p><blockquote><p>You have a very talented and skilled writting.</p></blockquote><p>Gmail has a little feed agreggator, I’ve got mine set up just to show quotes of the day. After a few days offline getting this today seems rather apt :</p><blockquote><p> I do not fear computers. I fear the lack of them.</p></blockquote><p>- Isaac Asimov</p><p>Golly, <a href="http://planetrdf.com/">Planet RDF</a> has been busy. My Bloglines account is stuffed to the gills too, but that’ll have to wait a while. Mail first, then I need to see what happened re. <a href="http://structuredblogging.org">Structured Blogging</a>, then? Dunno that’s probably the rest of today spoken for, but I should make some notes re. Trento. Oh, and there’s cleaning out Eric the hedgehog’s apartment. Caro weighed him again, he now looks like Cartman, doubled in weight in the few weeks we’ve had him. Need to find out if that means he can/should hibernate. It is very, very cold here right now though.</p><p>Tabloid headline of the week : <em>Nobleman Eats Dogfood</em></p>
<p>Scríobh Conn (<a href="http3A2F2Fimeall.blogspot.com2F">An tImeall</a>) alt faoi Wiki Ireland d� hAoine seo caite…</p><blockquote><p>Eola� is ea John Breslin at� i mbun taighde �r ch�rsa� idirl�n in Ollscoil na h�ireann, Gaillimh. T� su�omh seolta aige a bhfuil s� de chusp�ir aige b�aloideas agus eolas �iti�il a bhaili� � phobail agus � cheantair uile na t�re. An m� bailitheoir b�aloidis agus saor HTML at� fostaithe aige chun an gaisce seo a chur i gcr�ch? N�l aon duine! An cuimhin libh an Wikipedia go rabhas ag tr�cht air an tseachtain seo chaite? Wiki is ea an su�omh nua seo chomh maith: Wiki Ireland, agus t� cead agus cuireadh tugtha ag John do aon duine ar mian leo leathanaigh an tsu�mh a leas� agus cur leis an st�r eolais at� d� th�g�il ann.</p><p>Beart fadradharcach, nu�lach, sh�lfe�? Bheadh an ceart agat. N� nach ionadh gur thug Cumann Idirl�n na h�ireann duais “Aisling an Idirl�n” do John i m� na Samhna, ar mhaithe leis seo agus le obair eile at� d�anta aige.</p></blockquote><p><a href="http://imeall.blogspot.com/2005/12/62-ag-freastal-ar-dh-threo.html">Ar lean…</a></p>