IRC log of mmsem on 2006-12-08

Timestamps are in UTC.

11:45:23 [RRSAgent]
RRSAgent has joined #mmsem
11:45:23 [RRSAgent]
logging to
11:45:53 [Jacco]
11:46:11 [Jacco]
rrsagent, make logs public
11:46:19 [Jacco]
rrsagent, draft minutes
11:46:19 [RRSAgent]
I have made the request to generate Jacco
11:47:58 [Jacco]
zakim, list conferences
11:47:58 [Zakim]
I see no active conferences and none scheduled near this time
11:48:10 [raphael]
raphael has joined #mmsem
11:52:27 [Jacco]
Jacco has changed the topic to: MMSEM Athens F2F Meeting
11:53:05 [Jacco]
Meeting: MMSEM Athens F2F Meeting, December 8-9 2006
11:53:11 [Jacco]
rrsagent, draft minutes
11:53:11 [RRSAgent]
I have made the request to generate Jacco
12:02:29 [raphael]
Chair: Jeff, Raphael
12:09:01 [raphael]
Good afternoon everybody
12:10:58 [IoannisPratikakis]
IoannisPratikakis has joined #mmsem
12:11:33 [mhausenblas]
mhausenblas has joined #mmsem
12:13:35 [raphael]
Present: Giorgos Stamou, Susanne Boll, Sofia, Stamatia, Chrisa Tsinsaraki, Paola, Oscar, Eric, Tobias, Michael, Thierry, Tomas, Ioannis, Raphael, Jeff, Vassilis
12:14:34 [raphael]
Jeff open the meeting
12:14:45 [raphael]
We have problems to connect to Zakim ...
12:14:51 [raphael]
Scribe: Raphael
12:15:03 [raphael]
rrsagent, draft minutes
12:15:03 [RRSAgent]
I have made the request to generate raphael
12:15:24 [ocelma]
ocelma has joined #mmsem
12:17:11 [raphael]
Jeff talk about the possibilities in the future, and mention the opportunity of having a WG
12:17:40 [raphael]
... smooth transition between an XG to a WG
12:18:54 [raphael]
... will be a topic of discussion tomorrow
12:19:08 [raphael]
... the previous XG (content labeling) is trying to become a WG
12:22:49 [Jacco]
Jacco has joined #mmsem
12:24:03 [Jacco]
rrsagent, draft minutes
12:24:03 [RRSAgent]
I have made the request to generate Jacco
12:24:37 [raphael]
I'm trying to scribe
12:25:36 [raphael]
Topic: 1. Music Recommendation Use Case
12:25:43 [raphael]
Oscar Celma is
12:25:50 [raphael]
an official new invited expert
12:26:42 [raphael]
Introduction: personal music collections have grown
12:28:00 [raphael]
Music information Place: Audio (music recording), Text (lyrics), Video (video clips, CD covers)
12:28:12 [Paola]
Paola has joined #mmsem
12:28:46 [raphael]
Song metadata includes editorial, acoustic and cultural information
12:29:30 [raphael]
Users are represented by their profiles including demographic, geographic and psychographic information
12:30:27 [raphael]
Use case
12:32:06 [raphael]
Metadata initiatives for describing Audio Content
12:33:12 [raphael]
ID3 v2, OGG, APE v2, MusicBrainz 2.1 (RDF data not yet supported)
12:33:23 [raphael]
see vocabularies page
12:33:48 [raphael]
ID3, OGG and APE are tag based (free attribute value pairs)
12:34:20 [raphael]
All these approaches focus on editorial information
12:36:01 [raphael]
Audio Ontologies (in OWL)
12:36:18 [raphael]
Kanzaki, Music Production, Music Recommendation
12:36:21 [raphael]
See vocabularies
12:37:45 [raphael]
OGG Vorbis example
12:37:57 [raphael]
Interoperability problems and solutions:
12:39:07 [raphael]
... Two general approaches: 1/ integration at RDF level (song, artist instances) ; 2/ ontology mapping ?
12:42:50 [raphael]
... but there is no really ontologies
12:43:01 [raphael]
Concrete example, see code in the wiki
12:45:13 [raphael]
Idea: take the commonalities between all the proposed tags by these schema (ID3, OGG, etc.)
12:45:47 [raphael]
Vassilis: for which reasons the ontology has been made ?
12:46:37 [raphael]
ID3 tags are inside the audio file
12:48:04 [raphael]
Jeff: is there some conflicting name between the various schemas ?
12:49:39 [raphael]
Advantages: possible scenarios ?
12:50:06 [raphael]
Create personalized playlists ... exploit song metadata (RDF), plous user profile
12:50:54 [mhausenblas]
12:51:23 [raphael]
Upcoming gigs (user FOAF profile, geonames, RDF data about the venues and events
12:53:15 [EM]
EM has joined #mmsem
12:53:19 [raphael]
Simplify navigation in music collections through personalized views
12:53:54 [raphael]
integrate RDF music data into /facet
12:54:45 [raphael]
EM is Erik Mannens
12:55:04 [raphael]
Semi-automatic annotation of music
12:55:12 [raphael]
(link with the tagging use case)
12:56:50 [raphael]
Semantic podcast
12:58:42 [raphael]
Mention a wikilyrics that could have a webservice for getting the lyricsof a song
13:00:14 [raphael]
Paola: relation between a consumer and buisness; why not B2B ?
13:04:18 [raphael]
Oscar: it is possible
13:04:30 [raphael]
Future Plans:
13:05:33 [raphael]
search for commonalities between the various schema, build under the Music Recommendation ontology
13:05:49 [raphael]
Do a ID3 to RDF translation
13:06:38 [raphael]
ACTION: Oscar and Jacco to provide a first implementation by Christmas
13:06:58 [raphael]
ACTION: Paola to provide a possible B2B scenario
13:07:31 [raphael]
Topic: 2. News Use Case (Thierry Declerk)
13:09:03 [ocelma]
ocelma has joined #mmsem
13:11:07 [mhausenblas]
ScribeNick: mhausenblas
13:12:32 [mhausenblas]
Raphael: General introduction, NewsML, XBRL. UC will focus on financial domain.
13:13:18 [mhausenblas]
Thierry explains the 'Metadata Labels
13:15:11 [mhausenblas]
... TITLE tag contains a list (even sentence) of keywords
13:16:13 [mhausenblas]
... not sure if keywords stem from a controlled vocabulary
13:17:19 [raphael]
rrsagent, draft minutes
13:17:19 [RRSAgent]
I have made the request to generate raphael
13:18:34 [mhausenblas]
mhausenblas has joined #mmsem
13:19:03 [raphael]
Translating XBRL into Description Logic
13:20:08 [raphael]
Information about companie, profit and loss, etc.
13:21:36 [mhausenblas]
ScribeNick: mhausenblas
13:23:48 [raphael]
Possible application: expose NewsML information in RDF
13:24:00 [mhausenblas]
Raphael: Combine (lightweight) NewsML with XBRL descriptions
13:24:05 [raphael]
link them with RDF version of XBRL and DownJonesML
13:24:05 [raphael]
B2B scenario
13:24:25 [raphael]
Challenge: merging all the data
13:25:05 [mhausenblas]
Jeff: Concrete Examples?
13:26:00 [mhausenblas]
Raphael: Hopefully 01/2007 real world data should be available
13:26:35 [mhausenblas]
... Chirs HW is in touch with DowJones
13:27:45 [mhausenblas]
13:30:41 [mhausenblas]
Raphael: Any contribution is welcome. Michael will try to get some real world examples from NM2
13:33:02 [mhausenblas]
ACTION: Raphael, Therry, and Chris to sync UC Wiki description with their new focus (financial domain)
13:35:08 [mhausenblas]
rrsagent, draft minutes
13:35:08 [RRSAgent]
I have made the request to generate mhausenblas
13:36:00 [mhausenblas]
Paola: How about using Reuters test corpus?
13:36:59 [mhausenblas]
ACTION: Paola to supply XG with reference to corpus and usage license till next week.
13:37:49 [mhausenblas]
rrsagent, draft minutes
13:37:49 [RRSAgent]
I have made the request to generate mhausenblas
13:40:47 [raphael]
raphael has joined #mmsem
13:42:23 [chalaschek]
chalaschek has joined #MMSEM
13:53:48 [Zakim]
Zakim has left #mmsem
14:04:44 [mhausenblas]
mhausenblas has joined #mmsem
14:06:36 [raphael]
raphael has joined #mmsem
14:06:53 [raphael]
Back after the coffee break
14:07:47 [raphael]
rrsagent, draft minutes
14:07:47 [RRSAgent]
I have made the request to generate raphael
14:07:54 [EM]
EM has joined #mmsem
14:08:07 [raphael]
Topic: 3. MPEG-7 Use Case (Stamatia)
14:08:41 [raphael]
rrsagent, draft minutes
14:08:41 [RRSAgent]
I have made the request to generate raphael
14:08:51 [ocelma]
ocelma has joined #mmsem
14:11:51 [mhausenb]
mhausenb has joined #mmsem
14:12:46 [IoannisPratikakis]
IoannisPratikakis has joined #mmsem
14:13:39 [raphael]
Chrisa Tsinaraki will join this use case
14:15:04 [raphael]
Motivation: MPEG-7 is a set of XML schemas
14:15:24 [raphael]
... lack of explicit and precise semantics
14:15:44 [raphael]
... effort to attach formal semantics to mpeg7 (ontologies)
14:16:19 [raphael]
Example 1: semantically identically metadata can be represented in multiple ways
14:17:13 [raphael]
Example 2: image representation in itself as a multimedia entity can be performed in multiple ways was well
14:17:21 [raphael]
14:18:03 [raphael]
Example 3: a video segment can be a frame, a shot, a program
14:18:44 [raphael]
Example 4: loose meaning of decomposition: not part/whole semantics
14:18:56 [raphael]
14:19:35 [raphael]
Hunter: initially in DAML+OIL and then RDFS and OWL
14:19:44 [raphael]
... link to the ABC core model
14:20:15 [raphael]
... decomposition is captured using subclass constructs
14:21:01 [raphael]
raphel: actually, the result of the decomposition, the segment !!!
14:21:28 [raphael]
no formal axioms: an image cannot be decomposed temporally
14:21:51 [raphael]
SMARTWeb approach: same problem, no such axioms
14:22:39 [raphael]
Tsinaraki: translation of MPEG-7 MDS into OWL
14:24:51 [raphael]
Example whith the variations between Tsinaraki and Hunter ontologies
14:25:41 [raphael]
Possible solutions: 1/ mappings between the various ontologies ; 2/ buil a core ontology ? ; 3/ formalize the profiles
14:25:49 [raphael]
14:26:39 [raphael]
rrsagent, draft minutes
14:26:39 [RRSAgent]
I have made the request to generate raphael
14:27:49 [raphael]
Scribenick: raphael
14:30:00 [raphael]
Chrisa presents the DS-MIRF framework
14:30:49 [mhausenblas]
mhausenblas has joined #mmsem
14:32:42 [raphael]
Transformations from OWL/RDF to MPEG-7/21 in both ways
14:33:21 [raphael]
use of a reconstruction ontology: introducing the ordering of the elements, the default values
14:35:45 [raphael]
the solution of the use case could be based on Chrisa approach
14:37:01 [ocelma]
+1 too!
14:38:03 [raphael]
Sophia and Chrisa will join this use case
14:40:32 [raphael]
PB: nobody knows a use of some MPEG-7 ontologies ...
14:40:53 [raphael]
Raphael: why using these two ontologies ?
14:41:24 [raphael]
Vassilis: we do not focus on these two ontologies but rather on the mechanism for bridging multimedia ontologies
14:42:38 [raphael]
Michael: what this use case is about ?
14:42:59 [raphael]
Vassilis: there is various MPEG-7 ontologies
14:45:37 [raphael]
Oscar: propose to add a new document focusing on MPEG-7
14:45:50 [raphael]
... taking the part in the vocabularies deliverable
14:48:03 [raphael]
Raphael: why not making a new document ?
14:48:37 [ocelma]
that's what I meant! :-)
14:49:36 [mhausenblas]
+1 , volunteer for creating it ...
14:50:57 [ocelma]
+1 to contribute it (with Roberto Garcia, I hope!), once its created.
14:51:44 [raphael]
+1 to contribute too
14:52:45 [raphael]
Yannis: establish mappings between this multimedia ontologies and the existing vocabularies (FOAF, DC, ...)
14:54:28 [raphael]
... would be more lessons learned that can be further used rather than an example of a use case
14:59:18 [raphael]
Proposal: to take off MPEG-7 from the use case list and make a new standalone document
14:59:34 [mhausenblas]
14:59:45 [raphael]
this document will focus on MPEG-7, the various ontologies, and their interoperability problems
15:01:12 [raphael]
ACTION: Vassilis, Raphael, Oscar, Michael, Stamatia to set up a new Deliverabe
15:01:29 [raphael]
15:02:32 [raphael]
Thierry and all DFKI will join this deliverable
15:03:18 [raphael]
Topic: 4. Photo Use Case
15:03:28 [mhausenblas]
s/all/Paul and Massimo at
15:05:13 [raphael]
Susanne Boll presents the use case
15:05:35 [raphael]
rrsagent, draft minutes
15:05:35 [RRSAgent]
I have made the request to generate raphael
15:06:25 [raphael]
Photos: today (pixel-based + additional information)
15:06:37 [raphael]
tomorrow: capture and derive semantics annotation
15:07:16 [raphael]
What we want to know about photos: who, when, where, how long, weather, place, location, surrounding, company, quality, what I did, moud
15:08:15 [raphael]
Is tagging enough ?
15:08:57 [raphael]
there is not unique reasonable description scheme for photos (EXIF, photoRDF, IPTC)
15:09:44 [raphael]
photo undergo cross-platform distribution search
15:12:22 [raphael]
need for low-level features
15:13:12 [raphael]
Yiannis: MPEG-A about photo collections
15:13:43 [mhausenblas]
15:14:18 [raphael]
need for context and content
15:15:05 [ocelma]
(similar as Music UC)
15:15:48 [ocelma]
ocelma has joined #mmsem
15:16:37 [raphael]
Existing standards: EXIF, IPTC, PhotoRDF, FOAF, DC, MPEG-7
15:16:52 [raphael]
15:17:28 [raphael]
15:18:01 [raphael]
Interoperability problems:
15:18:18 [raphael]
... metadata preservation and enrichment
15:18:30 [raphael]
... centralized vs decentralized storage
15:18:52 [raphael]
... lacking lean but comprehensive standard
15:24:55 [raphael]
What could be the solution ?
15:27:30 [raphael]
Face recognition is really important, as well as Event (using potentially contextual information)
15:33:54 [raphael]
Propose to have a standard way to include in the photo a link towards the description
15:34:11 [raphael]
... so that photo application system can use the metadata to do ... cool stuff :-)
15:34:17 [raphael]
What kind of metadata ?
15:35:25 [raphael]
Focus on EXIF, a standard for visual features, a vocabulary for the content ...
15:35:40 [raphael]
Raphael: will you use RDF at all ?
15:35:51 [raphael]
Susanne: I'm not against, but not big fan !
15:36:00 [raphael]
... my concern is about the size of the description
15:37:50 [raphael]
Erik will join this use case
15:40:27 [raphael]
rrsagent, draft minutes
15:40:27 [RRSAgent]
I have made the request to generate raphael
15:42:13 [raphael]
ACTION: Susanne to look again in the various systems and standards and identify 3-4 standards that together could be the solution
15:43:39 [raphael]
Topic: 5. Multimedia Authoring Use Case
15:44:13 [raphael]
15:45:17 [raphael]
Example of a SMIL document
15:46:05 [raphael]
Problem: you use metadata in the select and assemble processes, but there is a semantics barriers
15:46:25 [raphael]
... as soon as you transform it in SMIL or in Flash ... you loose the metadata !
15:48:34 [raphael]
What is lost: structural information is hidden in the format of the presentation, emergent semantics when authoring
15:49:09 [raphael]
What is needed: making discourse and structural information explicit for search and access to documents and fragments
15:49:49 [raphael]
Existing standards: LOM (Learning Objects metadata) and SMIL metadata module
15:50:20 [raphael]
Multimedia documents hide almost everything ... no description for search
15:50:35 [raphael]
... an annotation problem, as well as an interoperability
15:57:47 [raphael]
Proposal: is this use case a clear open issue ?
15:59:08 [raphael]
... no solutions seem exist
15:59:17 [raphael]
... we all clearly see the problems
15:59:32 [raphael]
... but not the solutions
15:59:49 [raphael]
... susanne will continue to edit her thoughts on the wiki
16:00:30 [raphael]
rrsagent, draft minutes
16:00:30 [RRSAgent]
I have made the request to generate raphael
16:05:34 [raphael]
ACTION: Erik to remind Susanne about what is this standard your group has developped !
16:12:32 [raphael]
Topic: 6. Semantic Media Retrieval Use Case
16:13:38 [raphael]
Tobias will join the previous use case (Multimedia Authoring)
16:14:55 [raphael]
Ioannis Pratikakis
16:15:05 [raphael]
(From the Analysis Perspective)
16:15:52 [raphael]
Deal with semantic interoperability for automatic media interpretation / understanding
16:16:06 [raphael]
linking of low-level features to high level semantics
16:17:45 [raphael]
Example 1: lack of semantics in low-level descriptors (single media)
16:18:01 [raphael]
Example 2: interchange of semantics between media (multiple media)
16:18:53 [raphael]
This use case related to interchange of low-level descriptors for semantic interpretation
16:20:01 [raphael]
... exchanging classifiers is more interesting that exchanging descriptors
16:20:36 [raphael]
... MPEG-7 is useful for content-based retrieval
16:20:47 [raphael]
... but not for any semantic access
16:24:21 [raphael]
Interoperability problems
16:25:53 [raphael]
A user might want to exchange class descriptions between an imaging device, a set-top box, etc ... the process should be exchanged
16:26:05 [raphael]
16:26:21 [raphael]
exchange classifier parameters, and class descriptions
16:28:31 [raphael]
sketch of solution: multimedia core ontology, media descriptor ontology, classifier ontology, domain ontology
16:29:52 [raphael]
Raphael: sounds really close to the algorithm use case
16:32:10 [raphael]
Example contained in the use cases
16:32:24 [raphael]
rrsagent, draft minutes
16:32:24 [RRSAgent]
I have made the request to generate raphael
16:41:11 [raphael]
Vassilis: sounds like more a fusion problem that an interoperability problems
16:43:59 [raphael]
... will be discussed more tomorrow with the algorithm use case
16:44:07 [raphael]
... seems to have a lot of overlap
16:45:33 [raphael]
That close the meeting
16:48:51 [raphael]
Meeting starts tomorrow at 9:30 Greek time !!!!
16:48:54 [raphael]
See you tomorrow
16:49:00 [raphael]
rrsagent, draft minutes
16:49:00 [RRSAgent]
I have made the request to generate raphael