Back from BlogTalk Reloaded
Back from BlogTalk Reloaded a few days ago, where I enjoyed a lot of interesting talks and presentations[1] about social software and weblogging, especially :
- danah boyd’s keynote and especially remarks about relations between users and the way social networking websites are currently developped (hacks, reaction to users demands …);
- talks by Lee Bryant / Headshift and by Rod Smith / IBM about social software and use of open APIs and “Web 2.0″ technologies as AJAX, Google Maps … in corporate contexts;
- social aspects of blogging; even if I attended the conference mostly for its technical and corporate talks, it as a good opportunity to get different views about its impacts;
- monday evening meeting, and discussions with zeroK, Ton Zijlstra, bazaar guys, and more. A nice way to know more about activites and interests of everyone;
- Matt Mullenweng’s keynote. I was wondering since the program was published what what this “Jazz” about: new API ? Framework ? … No, simply music. Nice jazzmen quotes and pictures to describe the life of a software project - and maybe life itself ? The kind of keynote to relax after lunch !
Regarding my talk - apologizes for the english 
- what I felt is that Semantic Web still fears people (especially using and maintaining ontologies / knowledge bases), but I hope attendees were interested in the approach. I also got this feeling during SIOC presentation (.ppt) by Uldis and John, especially regarding the questions about microformats and SIOC differences and advantages. John wrote some ideas about creating links between both communities.
So, thanks again to Thomas, Jan and BlogTalk organizers for these two days.
Just a few suggestions:
- let speakers choose between different formats for their talks. While some presentations are indeed more appropriate for a 10 minutes talk and a debate, others are better with a more formal talk and Q/A way;
- proceedings. Even if a book is planned, it would be nice to get the papers at the beginning of the conference;
- shirts, please provice blogtalk shirts for us geeks next year
Finally, for those that didn’t attend, there’s a lot of stuff to look at on BlogTalk wiki, Technorati, Flickr …
Notes
[1] All talks where video-recorded and uploaded a few minutes after, and are still available on the program page.
BlogTalk Reloaded
I’ll flight to Vienna tomorrow morning to attend BlogTalk Reloaded.
My talk about “Folksonomies, Ontologies and Corporate Blogging” is planned to be monday at noon, while Uldis should present the one dedicated to SIOC (co-writen with John[1] and myself) on thurdsay morning.
There should be a lot of interesting talks during these two days, I’m especially waiting for the ones about social software in companies, presentation of the Google-Data API, but also about social aspects of blogging. Well, I also hope to meet new people, so if you’re frequently reading this blog, just come to say hello.
Notes
[1] Who recently made different SIOC Guides
Tags: blog, blogtalkreloaded, conference
New query interfaces for browsing SIOC data
I’ve recently added new queries (and associated interfaces) to my triple-store-based SIOC browser[1], in order to present what SIOC can be used for.
The first one is quite similar to the tagcloud feature (which creates a tagcloud of topics, as Technorati does but using sioc:topic property to get topics / categories): it displays a 3D pie of the ten most popular topics, using the PHP/SWF Charts API.
The second one uses the user aspect of SIOC. It has recently been decided to use both foaf:Person and sioc:User to represent creator of a sioc:Post when exporting data, while using only FOAF for unregistered users (i.e. comments on most weblog engines). This means that every Post (whether is it an original post or a reply) is linked to a foaf:User. Then, we can create mappings and relationships between posts and users, eventually smushing data using foaf:mbox_sha1sum if users use different nicknames but same e-mail in different blogs. So, this second query displays relationships beetween users within a set of decentralized SIOC-ed blogs that have been put in the same store. Two users are connected as soon as one replied to another, providing a way to discover people you’ve got in common (sometimes without knowing it !). The Prefuse API have been used to draw the network.
This FOAF/SIOC feature that links users to post have been recently introduced in exporters (WordPress, DotClear, b2evo) and in the PHP API. So, if you use an old version of one of this exporter, update it! These new versions also provide mappings between SIOC, FOAF and DC, and removed depracates some properties (see latest ontology specs). At the moment, I only store blogs that use the latest exporters in the store, so that all of them will be taken into accout in browsing interfaces, as the SPARQL queries refer to latest specs.
Finally, some of these exporting and browsing features will be presented in an upcoming talk at BlogTalk Reloaded, with the paper “SIOC Browser - Towards a Richer Blog Browsing Experience” co-written with John Breslin and Uldis Bojars. I’ll also be involved in a talk for the paper “Folksonomies, Ontologies and Corporate Blogging”, co-written with Philippe Laublet and Jean-David Sta, introducing an approach - using SIOC - that mix folksonomies and ontologies I dealt with in my PhD work.
So, once again, don’t hesitate to use SIOC exporters on your blog / forums if available (or write one with the API !) and get listed here !
Tags: blogtalkreloaded, rdf, sioc, socialnetwork, sparql
-
Current highlights
