See also: IRC log
<scribe> scribe: dF
<daveL> http://www.w3.org/2012/05/10-mlw-lt-minutes.html
<tadej> is this via gotomeeting?
Dave: Any issues with the
minutes?
... any additions to the agenda?
We are using the automated agenda here
<Yves_> Chair: DaveL
<Yves_> Scribe: dF
Yves, does the capitalization matter?
<Yves_> not sure. don't think so
I was setting chair and scribe few lines above.. just with lower case..
Dave: any aditional issues?
We start with Issue 14, Process Trigger
We discussed action 79, which was to consolidate this data category
cache was reintroduced, as we became aware that it is not part of the class
<daveL> http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2012May/0155.html
Dave: Pedro, can you wlak us through email on action 35?
Pedro: It is a simple expample, this metdata is to generated by the system
source and target are being compared for modifications
Private data such as names do not get cahced
cached
news should not be cached due to high amount of change
data with personal infomation is not cached
disclaimer [boiler plate type of information] is cached, as it is not likely to change
This type of metadata can help chosing the right real time MT system
scenario is different providers for different target languages
Dave: Do you have a text version of this example?
Pedro: Yes
Dave: I will go through it
tonight
... Close action 85? Yes
Close Issue [13 or 14] on cache with the action
Pedro: OK
It is Issue 13
Cache data category needed and how? (ISSUE-13 cache)
Pedro and Dave discuss action 79
ACTION-79 pending review Consider consolidation of status-related data categories and process trigger
we discussed source language
you want to add now the target language
Dave: We will keep the issue open
I will answer to the e-mail
DaveL to answer Pedro's e-mail on Issue 14 - process trigger
Pedro: souce and target lnaguge and maybe intermediate language
also priority
readiness for localization is also important
Dave: Discussion on API?
Is your thinking of appropriate attributes driven by the API discussion?
Pedro: No, it is a seprate issue
Dave: I understand TAUS are making something about an MT API
Arle: Jaap tries to revive the old translation webservices
Dave: My Masters student tries to do that with exposed categories through CMIS API
This tsudent will be in the June workshop
Dave: Defining API might be out of scope, but it is a concrete way of implemnting data categories
Yves, it is importnat to have at least two implementations
Dave to contact Moritz on API progress
<scribe> ACTION: Dave, to contact Moritz on API progress [recorded in http://www.w3.org/2012/05/17-mlw-lt-minutes.html#action01]
<trackbot> Sorry, couldn't find user - Dave,
<scribe> ACTION: DaveL, to contact Moritz on API progress [recorded in http://www.w3.org/2012/05/17-mlw-lt-minutes.html#action02]
<trackbot> Sorry, couldn't find user - DaveL,
<scribe> ACTION: daveL, to contact Moritz on API progress [recorded in http://www.w3.org/2012/05/17-mlw-lt-minutes.html#action03]
<trackbot> Sorry, couldn't find user - daveL,
Tadej on Action 80
How to solve issue of something not being in the domain
I would go with NERD
and would recommend mandating a single ontology
Dave: It will be difficult to find a single ontology, we should rather point to ontologies
We can recommned NERD in the informative section
Tadej: The ontologies do not operate our domains. This is for Issue 3
Thomas from Lucey answered me
MT disambiguation should be changed into general disambiguation
semantic selector
disambiguation data category
Tomas agreed that these reduce to concept pouinting
named entity has additional cetagories
terminology may have pointers
Text analytics anntotation: which agent created and confidence score
Should we consolidate confidence in text analytics and MT?
Dave: Any ideas? We do not have Declan
Pedro: I think they are different
Dave: Separate point, lot of cases where they will be used seprately
Pedro: It will be hard to provide reference ontology
In w3c there is a working group on ontologies
Dave: the NERD guys Tadej talked about work on lingustic ontologies
They will be in the June workshop
We will have opportunity to talk with people like Sebastian Hellmann
Tadej: NERD Is used in practice and for social
Dave: discussion in linked open data, different to the traditional RDF approach
People might be reacting agianst RDF
Pedro: we made clear differnce between reference and denotation
Tadej: examples?
Pedro: It may be critical for MT
Semantic selectors allow to do semantic disambiguation
mouse, refernce biology, or computong
This is also using linguistic ontologies and vocabularies
Tadej: I thought this will happen inside a compnate, nad intermediate steps won't be documented
Pedro: Euro Wordnet
to identify different meanings, we used domain ontologies and linguistic ontologies
Parsing analysis may be as useful as a domain category
Tadej: Exmaple of a semantic selector? Living creature? Countable?
Pedro: Most useful for versbs, as they are semantically complicated
<scribe> ACTION: Pedro, to go and find examples [recorded in http://www.w3.org/2012/05/17-mlw-lt-minutes.html#action04]
<trackbot> Sorry, couldn't find user - Pedro,
Pedro: I will send an example of to concept ontology
There are documented semantic features
top concept ontology
+Pedro
Pedro is on GoTo meeting :-)
Tadej: There is point in adding semantic selcetors, we need examples
Buit the issue is that we do not have producers and consumers for this type of categories
This happnes usually inside the black box
Dave: Isseu on Xpath version
(ISSUE-7 XPath Version)
Yves: I think we had resolution on this
<scribe> ACTION: daveL to remind Jirka to summarize [recorded in http://www.w3.org/2012/05/17-mlw-lt-minutes.html#action05]
<trackbot> Created ACTION-92 - Remind Jirka to summarize [on David Lewis - due 2012-05-24].
Dave: Action items due next week
I hold action on consolidating of author, revisionAgent and translationAgent
Dave is going to open Issue 22 on the action 81
Dave; Lot of people on ELREC next week
Dave: Action 34, Locworld is moving forward
Arle: This iwll need to be addressed in the PC meetings
This is scribe.perl Revision: 1.136 of Date: 2011/05/12 12:01:43 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Found Scribe: dF Inferring ScribeNick: dF Found Scribe: dF Inferring ScribeNick: dF WARNING: No "Present: ... " found! Possibly Present: Arle Dave Milan Pedro Tadej Yves Yves_ dF daveL mihael trackbot You can indicate people for the Present list like this: <dbooth> Present: dbooth jonathan mary <dbooth> Present+ amy Got date from IRC log name: 17 May 2012 Guessing minutes URL: http://www.w3.org/2012/05/17-mlw-lt-minutes.html People with action items: dave davel pedro WARNING: Input appears to use implicit continuation lines. You may need the "-implicitContinuations" option.[End of scribe.perl diagnostic output]