08:11:29 <RRSAgent> RRSAgent has joined #mlwrome
08:11:29 <RRSAgent> logging to http://www.w3.org/2013/03/12-mlwrome-irc

08:11:39 <fsasaki> meeting: MLW workshop, rome, day 1

08:11:46 <fsasaki> agenda: http://www.multilingualweb.eu/documents/rome-workshop/rome-program
08:11:48 <fsasaki> chair: arle
08:11:52 <fsasaki> scribe: various
08:11:56 <Labra> Labra has joined #mlwrome

08:12:36 <fsasaki> topic: Welcome and Keynote
08:13:01 <fsasaki> Daniel Gustafson on behalf of FAO introducing the conference

08:13:24 <fsasaki> arle showing an example of internationalization needed
08:13:44 <fsasaki> arle: language and culture are not easy issues to solve
08:13:49 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
08:14:27 <Gudrun> Gudrun has joined #mlwrome
08:14:48 <fsasaki> arle: we will learn about things from other communities, which normally won't see
08:14:57 <Jirka> Jirka has joined #mlwrome
08:15:08 <fsasaki> .. this workshop series shows what actions across communities are needed
08:15:29 <fsasaki> .. the mlw workshop series drives the development of a community who takes up the challenge of the mlw
08:15:36 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
08:16:16 <fsasaki> arle: we also act as a katalysator for future projects that take up the challenge of the mlw
08:16:19 <fsasaki> .. e.g. European projects that work on the topics discussed here
08:16:33 <fsasaki> .. we want to improve the use of standards and BP
08:16:48 <fsasaki> .. and want to improve support of multilingual features in browser agents
08:17:00 <fsasaki> .. we have seen real engagement in this area to tackle the issues
08:17:05 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
08:17:12 <omstefanov> omstefanov has joined #mlwrome
08:18:19 <Clara> Clara has joined #mlwrome
08:19:29 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
08:19:40 <fsasaki> arle going through admin issues and the program

08:19:58 <fsasaki> s/Welcome and Keynote/Welcome and workshop intro/
08:20:00 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
08:23:40 <fsasaki> topic: Keynote: Innovations in Internationalization at Google
08:23:46 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
08:24:52 <fsasaki> mark: we will focus on some of the technologies at google and how we make our products more multilingual
08:24:55 <fsasaki> scribe: fsasaki
08:25:54 <fsasaki> mark: at google we have to deal with core localization - here we will talk about work which is above the core
08:26:06 <fsasaki> .. google is about search for text
08:26:28 <fsasaki> ... we take synonyms into account, but more recently we want to take entities into account
08:26:34 <fsasaki> .. "entity i18n"
08:26:48 <fsasaki> .. we do this e.g. to look at wikipedia to find out how entities look like
08:27:12 <fsasaki> .. english wikipedia is huge, so we did cross connections of wikipedia to find more about entities
08:27:35 <GordonD> GordonD has joined #mlwrome
08:27:35 <fsasaki> .. part of this is names, e.g. personal names, google+pages with more free form names
08:27:42 <fsasaki> .. and also URLs, which present their own problems
08:27:49 <fsasaki> .. some problems are related to security
08:28:00 <fsasaki> .. many characters are look alikes
08:28:30 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
08:28:30 <fsasaki> .. this creates opportunities for spoofing
08:28:45 <fsasaki> .. normalization of names: when are two names the same?
08:29:09 <fsasaki> .. that includes handling of inflection, how people want their names represented
08:29:36 <fsasaki> .. this involves issues related to semantics, encoding, formatting etc.
08:29:42 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
08:30:17 <fsasaki> mark: recently we worked on plurals and gender
08:32:21 <fsasaki> mark: plural and gender are tricky features to deal with
08:32:32 <fsasaki> .. currently we have patterns for numbers written as digits
08:32:42 <fsasaki> .. used in messages, units, contact numbers, etc.
08:33:08 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
08:33:27 <fsasaki> mark: now handing over to vladimiar who will take us through these areas
08:33:47 <fsasaki> vladimiar: we have a nice way to represent e.g. gender and number information across languages
08:33:55 <fsasaki> .. we want translators to handle with this properly
08:34:19 <fsasaki> .. we built a tool that would show the localization specialist what he needs to see
08:34:26 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
08:34:39 <fsasaki> vladimar: engineers write forms in English
08:34:55 <fsasaki> .. in our code we have ways to specifiy how gender and plural should be treated
08:35:19 <fsasaki> .. examples on how this works for various languages
08:35:25 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
08:35:44 <Andrea> Andrea has joined #mlwrome
08:35:44 <smokingred83> smokingred83 has joined #mlwrome
08:36:13 <fsasaki> vladimir: now phone number example
08:36:25 <fsasaki> .. operating system that runs on the phone
08:36:35 <fsasaki> .. for each phone number we want it to be unique
08:36:48 <fsasaki> .. that is, there is a country code, area code, digits to dial the number
08:37:09 <fsasaki> .. people in different countries don't want to think about +xx, they just want to write things as they like
08:37:21 <fsasaki> .. so we have an open source phone number library
08:37:53 <fsasaki> .. see http://code.google.com/p/libphonenumber/
08:38:07 <fsasaki> .. it handles parsing, formatting, canonicalization
08:38:11 <fsasaki> .. getting types and examples
08:38:14 <fsasaki> .. finding numbers in text
08:38:34 <fsasaki> vladimir: canonicalization is a hard problem
08:38:54 <fsasaki> .. google contact book has a flag that tells you the country of the number
08:39:12 <fsasaki> .. it allows you to fix the number if it's wrong, and it relies on above library
08:39:19 <fsasaki> .. geolocation of number is an issue
08:39:24 <fsasaki> .. in some places you cannot solve it
08:39:39 <fsasaki> .. in us there is a physical , teretorry designation
08:39:50 <fsasaki> .. in Europe that is different
08:39:59 <fsasaki> .. so the problem can't be solved in general
08:40:05 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
08:40:14 <fsasaki> vladimir: now about addresses
08:40:25 <fsasaki> .. e.g. for sending a check
08:41:10 <fsasaki> .. see library at http://code.google.com/p/libaddressinput/
08:41:13 <fsasaki> .. this is also open source
08:41:32 <fsasaki> .. allows for validation of regionsy, layout and basic validation
08:41:50 <fsasaki> .. to e.g. give a street address that is actually meaningful
08:42:28 <fsasaki> mark: "getting language settings wrong"  - an issue for users and also people inside Google
08:42:30 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
08:42:45 <fsasaki> .. was a hard problems to fix
08:43:04 <fsasaki> .. worse even for enterprises, since the language often is set but some administration
08:43:24 <fsasaki> .. we created "universal language settings"
08:43:34 <fsasaki> .. this allows people to set the language across languages
08:43:46 <fsasaki> .. sounds simple but rollout across products is hard
08:43:58 <fsasaki> .. allows for setting more than one language
08:44:18 <fsasaki> .. did some analysis of gmail - a fair amount of users speaks more than one language
08:44:22 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
08:44:40 <fsasaki> .. fallback is needed if preferred language is not available
08:45:02 <fsasaki> .. a use case is to serve better content in search scenarios
08:45:29 <fsasaki> .. with language settings the outcome is better than trying to gues the users language
08:45:36 <fsasaki> .. "60 language initative" at google
08:45:47 <fsasaki> .. we did a 40 language initiative
08:46:03 <DomJones> DomJones has joined #mlwrome
08:46:18 <Monica> Monica has joined #mlwrome
08:46:19 <fsasaki> .. we showed google internally how support for more languages helps to get more satisfied customers
08:46:29 <lbellido> lbellido has joined #mlwrome
08:46:37 <fsasaki> ..  that was an incentive google internally to have more language support
08:46:41 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
08:46:56 <fsasaki> .. now we are rolling out 60 languages in many products
08:47:41 <fsasaki> vladimir: important for people that they can interact with their device the way they want
08:48:25 <fsasaki> .. speech-text library important for that; our library for that now has support for 42 languages including accents / dialects in 46 countries
08:48:35 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
08:48:49 <ivan> ivan has joined #mlwrome
08:49:31 <fsasaki> vladimir: text input  - dedicated team is developing input methods for many languages
08:49:44 <fsasaki> .. input methods on android, native input methods to store on your device
08:49:51 <fsasaki> .. also cloud input methods
08:50:17 <fsasaki> .. librar(ies) for input http://www.google.com/inputtools/
08:50:47 <fsasaki> .. with dictinoaries and word frequency data we are trying to guess what people want to type
08:50:54 <fsasaki> .. which is helpful for many users
08:51:18 <fsasaki> .. another team at google creating fonts for all unicode scripts
08:51:30 <fsasaki> see http://code.google.com/p/noto/
08:51:35 <fsasaki> noto = no tofu
08:51:49 <fsasaki> vladimir: dealing with so many fonts and font data
08:52:06 <fsasaki> .. needs tools for reading and writing basic font tables, special bit maps glyphs
08:52:17 <fsasaki> .. allowing us to serve smaller subset of fonts
08:52:29 <fsasaki> .. we have a sfntly font library for that
08:52:56 <fsasaki> http://code.google.com/p/sfntly/
08:52:56 <fsasaki> vladimir: now google translate
08:53:03 <fsasaki> .. now support for 65+ languages
08:53:35 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
08:53:46 <fsasaki> .. if you use a chrome browser the browser can detect the content of the website and will offer translation
08:54:12 <fsasaki> .. we also allow you to submit feedback to make the translation better
08:54:20 <fsasaki> .. we are allowing people to access content on the web
08:54:28 <fsasaki> .. we do not shoot for specialized enginges
08:54:46 <fsasaki> vladimir: now google localization infrasttructure
08:54:57 <fsasaki> .. see http://translate.google.com/toolkit
08:55:16 <fsasaki> .. used now for all google localization
08:55:52 <fsasaki> .. important to have everything available under google control, so that changes can be rolled out everywhere easily
08:56:13 <fsasaki> .. the toolkit can also be used for outside users
08:56:26 <fsasaki> .. you can use translation tools, glossaries etc. for yourself
08:56:34 <fsasaki> s/tools/memories/
08:56:49 <fsasaki> vladimir: world of localization is not well standardized
08:56:55 <fsasaki> .. we created ARB format
08:57:00 <fsasaki> .. for web applications
08:57:34 <fsasaki> .. we created an easy to use JSON format to use at runtime to skin applications at runtime
08:57:52 <fsasaki> .. we also allow people to produce localized content
08:58:05 <fsasaki> .. example of youtube caption translation
08:58:10 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
08:58:36 <fsasaki> .. allows to take your captions from the video and ask your friends to translate it yourself
08:58:53 <smokingred83> smokingred83 has joined #mlwrome
08:59:15 <fsasaki> vladimir: next itegration even allows you to buy captions from a vendor
08:59:28 <nicoletta> nicoletta has joined #mlwrome
08:59:29 <fsasaki> .. idea is that people can access translation of their content on many different levels
08:59:37 <fsasaki> s/idea/general ideal/
08:59:43 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
08:59:54 <fsasaki> present: many, many, many, people

09:00:15 <fsasaki> topic: keynote q/a
09:00:21 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
09:00:42 <joomla> joomla has joined #mlwrome
09:01:06 <fsasaki> chaals: question on ARB format
09:01:27 <fsasaki> .. w3c standardized various XML formats related to that
09:01:36 <fsasaki> .. google didn't follow that path and choose the json way
09:01:42 <fsasaki> .. why did google took that path?
09:02:01 <fsasaki> mark: google for the format is to have something light
09:02:09 <fsasaki> . .that can be mapped to xml, but something simple
09:03:06 <fsasaki> tomas: do you follow language identification means?
09:03:10 <fsasaki> mark: we use bcp 47
09:03:19 <fsasaki> .. we do not follow language accept header in HTTP
09:03:33 <fsasaki> .. since that has never really been used in a consistent way
09:03:53 <fsasaki> tomas: sending in the header "this is my language preference" - how about that?
09:04:07 <fsasaki> mark: we cannot rely on people having set put the header right
09:04:21 <fsasaki> .. then people are also not depending on which machien they are working on
09:04:22 <Andrea> Andrea has joined #mlwrome
09:04:33 <fsasaki> .. things are handled just be the google (account) settings
09:04:50 <fsasaki> xyz: are you using timed text stanards?
09:05:00 <fsasaki> (= caption translation standard)
09:05:18 <fsasaki> vladimir: we support one standard here, not sure which one
09:05:29 <fsasaki> .. different standards for captions have different ways to convert from one to another
09:05:42 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
09:06:12 <fsasaki> christianL: you use quite a bit of natural language processing in your tools
09:06:20 <fsasaki> .. e.g. google translation in 65+ languages
09:06:42 <fsasaki> .. in some areas: are you still working with rule-based, that is not statistical methods?
09:07:04 <fsasaki> mark:google translate uses masses of bi-lingual data
09:07:13 <fsasaki> s/mark:google/mark: google
09:07:27 <fsasaki> mark: that doesn't work with languages that need ordering
09:07:36 <fsasaki> .. so there are now pre-ordering steps
09:07:46 <fsasaki> .. and syntactic analysis more and more being used
09:08:02 <fsasaki> .. advantage of data approach that easily you can accomodate new language pairs
09:08:07 <fsasaki> .. just with new data
09:08:23 <fsasaki> .. the  rule based approach was very labor extensive for us
09:08:28 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
09:08:44 <fsasaki> mark: google translate does not map into the most commercial languages
09:08:58 <fsasaki> .. but the languages that have most of the data
09:09:08 <fsasaki> tomas: do you tackle the data that you use in MT?
09:09:32 <fsasaki> .. if you have the choice between a smaller amount with good quality and large "dirty" data sets?
09:10:00 <fsasaki> mark: we take into account large data sets in standard format, but we also want to translate tweets
09:10:23 <fsasaki> .. also need to make sure we do not use data for training that has been MT translated already
09:10:44 <Monica> Monica has joined #mlwrome
09:10:49 <fsasaki> .. or example of date representation: sometimes auto-generated with CLDR, that then influences training
09:11:08 <RRSAgent> I'm logging. I don't understand 'drat minuts', fsasaki.  Try /msg RRSAgent help
09:11:11 <RRSAgent> I'm logging. I don't understand 'drat minutes', fsasaki.  Try /msg RRSAgent help
09:11:31 <fsasaki> richard: maybe explain what CLDR is
09:11:49 <fsasaki> .. and explain what mobile internet tool xyz does
09:11:58 <fsasaki> mark: cldr is a project for gathering localization data
09:12:15 <fsasaki> .. e.g. formatting of dates, times, numbers, collaction and sorting rules, currency formatting
09:12:23 <fsasaki> .. contact numbers
09:12:59 <fsasaki> s/xyz/halfbuzz/
09:13:13 <fsasaki> mark: halfbuzz is an open source project to handle complex scripts
09:13:26 <fsasaki> .. it is gradually developing to support more and more scripts in the world
09:13:30 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
09:14:05 <fsasaki> now coffeee break
09:14:10 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
09:20:44 <vp> vp has joined #mlwrome
09:38:04 <stocazzo> stocazzo has joined #mlwrome
09:38:49 <antonino_carella> antonino_carella has joined #mlwrome
09:39:13 <stocazzo> stocazzo has left #mlwrome
09:39:57 <lmatteis_> lmatteis_ has joined #mlwrome
09:40:04 <lmatteis_> hi all
09:40:07 <lmatteis_> what's up?
09:43:38 <lmatteis_> why nobody talk?
09:44:56 <tadej> tadej has joined #mlwrome
09:45:03 <laurent_oz> laurent_oz has joined #mlwrome
09:45:09 <lmatteis_> hi tadej and laurent_oz
09:46:10 <Monica> Monica has joined #mlwrome
09:49:09 <laurent_oz> hi everyone, Laurent from CSIRO/W3C Australia (Canberra). Not staying long but hoping to finsih a conversation with Felix Sasaki about NUTTAB http://www.foodstandards.gov.au/consumerinformation/nuttab2010/
09:49:53 <Andrea> Andrea has joined #mlwrome
09:50:15 <fsasaki> fsasaki has joined #mlwrome

09:50:22 <fsasaki> scribe: fsasaki
09:50:39 <fsasaki> topic: Going Global with Mobile App Development: Enabling the Connected Enterprise
09:50:44 <fsasaki> presentation by jan nelson
09:50:49 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
09:51:57 <fsasaki> jan: windows 8 phone toolkit
09:52:19 <fsasaki> .. visual studio IDE, provides pseudo languages for "in house" testing
09:52:25 <fsasaki> .. XLIFF support
09:52:36 <Clara> Clara has joined #mlwrome
09:52:39 <fsasaki> ... integration with MS trasnlator service
09:52:42 <fsasaki> .. via interent
09:52:55 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
09:53:00 <nicoletta> nicoletta has joined #mlwrome
09:53:13 <lbellido> lbellido has joined #mlwrome
09:53:34 <Arle> Arle has joined #mlwrome
09:53:43 <fsasaki> jan: windows phone 8  - you create a new project, binding to resources, then testing in various languages
09:53:54 <fsasaki> .. you test in many languages, apply resource changes across languages
09:54:06 <fsasaki> .. re-starting testing, ... takes a lot of work to ship in one language
09:54:23 <fsasaki> .. demo today in the toolkit: how we are trying to make the process easier
09:54:51 <fsasaki> .. will show how to add a new language, how to export to xilff, store the data in sky drive etc.
09:55:08 <fsasaki> .. so high level overview: how you can create a windows 8 app in various languages
09:55:40 <fsasaki> .. in the enterprise it is important: "bring your own device" scenario
09:56:02 <fsasaki> .. in the past that didn't work in the enterprise, due to security, protocols etc.
09:56:38 <fsasaki> ..  this tool supports that scenario, but also goes beyond the enterprise
09:56:41 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
09:57:39 <fsasaki> ian demoing the toolkit
09:57:54 <fsasaki> ian: resources file, an XLIFF file
09:58:03 <fsasaki> .. a specific locale for pseudo localization
09:58:42 <fsasaki> ian: now adding french as a new language
09:58:48 <fsasaki> .. there is a notation for a translator service
09:58:52 <antonino_carella> why so much attention on wp8? the most diffused platform is android
09:58:55 <fsasaki> .. this service adds more language pairs
09:59:00 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
09:59:29 <Monica> Monica has joined #mlwrome
09:59:35 <lmatteis> antonino_carella: true, isn't it one of the least used mobile operatying systems?
09:59:54 <lmatteis> and especially for w3c, it seems to me that open-source would be much more interesting
10:00:06 <lmatteis> things like Firefox OS are much more promising
10:00:12 <fsasaki> lmatteis, antonino, are you in the room? do you want to ask the questino during q/a session?
10:00:26 <antonino_carella> i'm not in the room
10:00:44 <fsasaki> besides: at this workshop we try not to evaluate company specific technologies - just show what it can do and what "standardization gaps" are here
10:00:50 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:01:17 <antonino_carella> yes of course
10:01:35 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:03:02 <lmatteis> fsasaki: which standards is the current presentor showing?
10:03:07 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:03:07 <lmatteis> also live stream has no more audio.
10:03:13 <fsasaki> lmatteis: XLIFF
10:03:24 <fsasaki> s/lmatteis/at lmatteis/
10:03:49 <fsasaki> the editor allows you to generate XLIFF (XML Localization Interchange file format)
10:03:56 <lmatteis> ok audio is back on live stream
10:04:26 <fsasaki> there was now audio, just video - is the video streaming working for you?
10:04:31 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:04:32 <lmatteis> now yes
10:04:40 <fsasaki> good, thanks
10:04:56 <fsasaki> ian continuing demo  - now loading resource file
10:05:03 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:05:14 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:05:53 <Arle> Arle has joined #mlwrome
10:06:26 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:07:42 <Andrea> Andrea has joined #mlwrome
10:08:17 <fsasaki> ian: in windows 8 you have language preference settings - you can add several ones and a fallback too
10:08:26 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:10:42 <fsasaki> ian: xliff support important, for working with many translator services
10:11:02 <fsasaki> .. we continue to work active in the XLIFF TC to assure that there will be interoperability
10:11:46 <fsasaki> topic: gavin brelstaff - Multilingual Mark-Up of Text-Audio Synchronization at a Word-by-Word Level – How HTML5 May Assist In-Browser Solutions
10:11:51 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:12:42 <lmatteis> fsasaki: there are no questions?
10:12:49 <lmatteis> (for the speaker)
10:12:59 <Monica> Monica has joined #mlwrome
10:13:24 <fsasaki> lmatteis, q/a will be at the end of the session
10:13:37 <lmatteis> ok

10:13:46 <fsasaki> see http://www.multilingualweb.eu/documents/rome-workshop/rome-program - noon - 12:15 "q/a"
10:14:32 <fsasaki> gavin: example of a movie - french movie with english subtitles
10:15:10 <fsasaki> .. example of closed captions in HTML5 - just put track, use timed text markup
10:15:11 <fsasaki> .. can be e.g. vtt, or srt standards
10:15:16 <fsasaki> .. it gives you line-by-line translation
10:15:19 <lmatteis> does youtube support this?
10:15:21 <fsasaki> .. supported by youtube
10:15:54 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:16:24 <fsasaki> example of english subtitles for italian clip on wikimedia commons
10:16:49 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:17:19 <fsasaki> gavin: difficult to type all numbers of timed text
10:17:25 <fsasaki> .. we will show later how that can be easier
10:18:08 <fsasaki> .. in pisa mlw workshop we gave a presentation
10:18:12 <fsasaki> .. of alignment of bilingual text
10:18:53 <fsasaki> gavin: example works on the desktop
10:19:29 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:20:09 <omstefanov> omstefanov has joined #mlwrome
10:20:13 <fsasaki> gavin: using javascript, html, jquery, and we can do a go through highlighting of the text
10:21:03 <fsasaki> gavin: semantics can spread on the whole part of the page, not always easy to extract
10:21:10 <fsasaki> .. somebody has put some timed text markup in
10:21:19 <fsasaki> .. we can say "we are here in the text"
10:21:25 <fsasaki> .. we don't make a video - we work on text
10:21:52 <lmatteis> hrm
10:21:57 <lmatteis> he didn't explain what this interface does
10:22:04 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:22:05 <lmatteis> i can guess, but it's not very clear
10:22:21 <fsasaki> gavin: a bit like karaoke
10:22:45 <fsasaki> .. you can hear it in the original language and follow in your language
10:22:58 <lmatteis> what you can?
10:23:06 <lmatteis> there was no intro on what that interface with text was
10:23:23 <fsasaki> demo of alignment across languages - left side italian, right side english, the highlighting of words moves with the audio
10:23:31 <Clara> Clara has joined #mlwrome
10:23:50 <fsasaki> gavin: we have a way to mark up semantic correspondance - green is direct correspondance
10:23:50 <lmatteis> ok... well good intuition i guess
10:24:06 <fsasaki> .. and there are other color codes for other types of equivalence
10:24:32 <fsasaki> .. HTML under the hood: audio element, no controls
10:24:40 <fsasaki> .. there are start and end + time identicators
10:24:47 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:25:00 <fsasaki> gavin: markup is reflecting "semantic segmentation"
10:25:36 <fsasaki> gavin: there is also an archived format, can be safed by server, an XML format specified in text encoding initative
10:26:02 <fsasaki> .. we do that for historical reasons - with JSON it might be archaic, but academics use it a lot
10:26:12 <fsasaki> .. the hard task is to deal with overlapping hierarchives
10:26:34 <fsasaki> .. it would be useful if some kind of timed text standard woudl address the overlap issue
10:26:45 <fsasaki> s/hierarchives/hierarchies/
10:26:50 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:28:37 <fsasaki> demo continuing
10:28:43 <Monica> Monica has joined #mlwrome
10:28:54 <fsasaki> gavin: visual interface to add cue points
10:29:01 <fsasaki> gavin: why are we doing this?
10:29:13 <fsasaki> .. aim to activate petic memory
10:29:43 <fsasaki> s/petic/poetic/
10:30:18 <fsasaki> .. complementary to external memory produced by europeana - we try to get the memomry inside heads of people
10:30:28 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:31:28 <fsasaki> talk finished
10:31:35 <fsasaki> thanks to galvin

10:32:03 <fsasaki> topic: Gábor Hojtsy - Multilingual Challenges from a Tool Developer's Perspective
10:32:08 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:33:17 <fsasaki> gabor: have a decade of experience on support of multiple languages in drupal
10:33:55 <fsasaki> .. drupal is mostly used as a CMS platform, used by various famous sites
10:34:15 <fsasaki> s/used by/examples of several/
10:34:20 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:35:23 <fsasaki> Gábor: first gettext .po format
10:35:30 <fsasaki> .. we use since it fulfils our needs
10:35:44 <fsasaki> Gábor: location information, plural format, message texts etc.
10:35:53 <fsasaki> .. we only use this as a transfortation format
10:35:58 <fsasaki> .. very simple, small
10:36:05 <fsasaki> .. we don't need to deal with gender issues
10:36:17 <fsasaki> Gábor: that's rarely an issue for drupal web sites
10:36:26 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:36:47 <fsasaki> Gábor: at drupal.org we have 20 000 modules hosted
10:37:14 <fsasaki> .. you need to have incentives for people to participate in open source efforts
10:37:34 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:37:47 <fsasaki> Gábor: we found that there is a lot of overlap between drupal projects
10:37:59 <fsasaki> .. when you use projects together you can share translations
10:38:18 <fsasaki> .. we built a sub model for bulding translations for drupal
10:38:29 <fsasaki> .. we found that encouraging micro contributions is helpful
10:38:35 <fsasaki> .. people can submit one translation each
10:38:41 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:38:58 <fsasaki> Gábor: we built a diffing tool etc. and people get fame by contributing more
10:39:34 <fsasaki> Gábor: configuation as a problematic area
10:39:43 <fsasaki> .. configuration in drupal can be edited by the user
10:40:58 <fsasaki> gabor: we use YAML for configuration
10:41:04 <fsasaki> .. we ship these with drupal itself
10:41:24 <fsasaki> .. people create their own configuration - people should ship the stuff as part of the software
10:41:37 <fsasaki> .. breaking it down to small pieces was very beneficial
10:42:02 <fsasaki> .. that is a kind of dual system that we needed to support
10:42:09 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:43:00 <fsasaki> gabor: getting acceptance for two different translation models is difficult - it will freak out developers
10:43:10 <fsasaki> .. you need to have one clean API
10:43:24 <fsasaki> .. we go with the 2nd model to identify pieces in the content
10:43:36 <fsasaki> .. for workflow support we have manual support for XLIFF
10:44:13 <fsasaki> .. we have vendors that build tools  - XLIFF tool, translation managemtn tool that has support for ITS and is demoed at the showcases
10:44:31 <fsasaki> s/managemtn/management tmgmt/
10:44:57 <fsasaki> (see more about ITS + tmgmt at http://drupal.org/sandbox/kfritsche/1908598 )
10:45:08 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki

10:45:35 <fsasaki> topic: Reinhard Schäler - Enabling the Global Conversation in Communities
10:45:39 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:46:27 <fsasaki> reinhard: a confession - I'm not a developer - but "every saint has a past, and every sinner has a future" :)
10:46:36 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:47:13 <fsasaki> reinhard: rosetta foundation is about empowering language communities
10:47:45 <fsasaki> .. two-three years ago we found that this is also about a business model
10:48:29 <fsasaki> reinhard: you can reach 90% of the customers by just localizing in 50 languages
10:48:43 <fsasaki> .. that's not 90% of population, but 90% of customers
10:48:55 <fsasaki> .. story of bloomberg storyful
10:49:17 <fsasaki> .. a startup. It tells you looking at tweets what effects stock prices
10:49:27 <fsasaki> .. access to information and knowledge is crucial for money
10:49:34 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:49:48 <fsasaki> .. "social localzation": non market localization
10:49:57 <fsasaki> .. localization for world piece - all the good things you can do
10:50:23 <fsasaki> .. with this you will reach 70% of citizens - not customers, but citizens
10:50:29 <fsasaki> .. many organizations active in this space
10:50:41 <fsasaki> .. so situation here: no business case, but lots of activity
10:50:51 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:51:12 <fsasaki> .. problem is how to connect content with volunteers that want to help?
10:51:24 <fsasaki> .. the answer we got was: why not use what we have got already?
10:51:50 <fsasaki> .. all the stuff we have works well in a commerical setting
10:51:57 <fsasaki> .. but often it is closed and working in silos
10:52:20 <fsasaki> .. with "705 of population" scenario anything can happen
10:52:38 <fsasaki> .. what we need is something open, configurable, standards based
10:52:45 <fsasaki> .. that is where XLIFF and w3c standards come in
10:53:02 <fsasaki> ... SOLAS is a localization architecture
10:53:10 <fsasaki> ..  I will talk about productivive part of SOLAS
10:53:30 <fsasaki> .. david filip and dave lewis have a demo here to show the interoperability in SOLAS productivity
10:53:43 <fsasaki> .. focus is on impact rather than commercial value
10:53:47 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:54:23 <ivan> ivan has joined #mlwrome
10:54:28 <fsasaki> reinhard: we are about to announce an open soure project about solas
10:54:43 <fsasaki> .. we have just signed a related licensing agreement with univ. of limerick
10:54:57 <fsasaki> .. we will open source it an announce that next week at the gala conference
10:55:25 <fsasaki> .. we see solas as the tool that will connect massive amount of content with massive number of people
10:55:37 <fsasaki> .. not only to produce wealth
10:55:48 <fsasaki> .. the project will run in trommons - translation commons
10:57:13 <fsasaki> various solas screenshots demonstrating the functionality
10:57:20 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:57:39 <fsasaki> reinhard: exampe of translating wikipedia article
10:57:47 <fsasaki> .. easy to organize with solas
10:58:05 <fsasaki> .. you can have tasks or sub tasks that volunteer transators can take up
10:58:10 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:58:39 <fsasaki> reinhard: matching interest of translators with specific tasks is the key
10:58:50 <fsasaki> .. that's why it is called solas match
10:58:58 <fsasaki> .. you find the preferred task that fits you
10:59:02 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
10:59:27 <fsasaki> .. we need to find bankers to support this

11:00:05 <fsasaki> topic: q/a for developers
11:00:19 <fsasaki> christian introducing q/a session
11:00:24 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
11:01:09 <fsasaki> ionannis: many different platforms
11:01:23 <fsasaki> .. for translation: google, microsoft, solas, ...
11:01:25 <fsasaki> .. drupal
11:01:28 <labra> labra has joined #mlwrome
11:01:37 <fsasaki> .. is there the danger to have too many platforms? Can this be united
11:01:49 <fsasaki> gavin: the standards based web is the platform of the future
11:01:58 <fsasaki> gabor: agree that this is a problem
11:02:15 <fsasaki> .. we had a problem in drupal too - if we build it for ourself it needs to be maintained, build the features
11:02:25 <fsasaki> .. if we cannot sustain interest it is a problem
11:02:48 <fsasaki> .. a different solution is to have a backend that we re-use and just have our interface on top of it
11:03:00 <fsasaki> reinhard: you need to integrate, standardize and strive for interop
11:03:02 <tgraham> tgraham has joined #mlwrome
11:03:08 <fsasaki> ioannis: is this happening?
11:03:13 <fsasaki> reinhard: in solas
11:03:30 <fsasaki> gavin: 5 years ago the browser wouldn't have give me technologoy to do what I demoed
11:03:39 <fsasaki> ian: you heard XLIFF, CLDR; Unicode, ...
11:03:46 <fsasaki> .. there are many standards we can work on
11:03:56 <fsasaki> .. think of the network stack
11:04:08 <fsasaki> .. this is similar: as we work on the multilingual stack
11:04:22 <fsasaki> .. we have to be careful on what becomes part of the standardization
11:04:34 <fsasaki> .. see e.g. the ITS shown at the workshop
11:04:45 <fsasaki> olaf: competition is important
11:04:49 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
11:05:12 <fsasaki> s/olaf/olaf-michael/
11:05:23 <fsasaki> .. as long as we have serious competition, standards are helpful
11:05:42 <fsasaki> .. if we don't have competition, things will not develop as fast as we'd like
11:06:00 <fsasaki> mark: at gabor - how do you manage your community
11:06:14 <fsasaki> .. if you get 10 different translation for the same thing for example?
11:06:21 <fsasaki> s/gabor/reinhard and gabor/
11:06:32 <AndChat496496> AndChat496496 has joined #mlwrome
11:06:32 <fsasaki> reinhard: we are trying to manage as little as possible
11:06:34 <fsasaki> .. and to trust
11:06:45 <fsasaki> .. that implies some monitoring and intervention
11:06:48 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
11:07:01 <fsasaki> reinhard: we don't have entry barriers
11:07:09 <fsasaki> .. people can take part and can contribute
11:07:19 <fsasaki> .. there are good examples in the open source community
11:07:22 <fsasaki> .. that this works
11:07:42 <fsasaki> reinhard: there are big questions around quality of translations
11:07:54 <fsasaki> .. but there are no of the shelf answers to this
11:07:58 <nwaltham> nwaltham has joined #mlwrome
11:08:05 <fsasaki> .. I think twitter is trying to work with 500 000 volunteers
11:08:11 <fsasaki> .. there is no silver bullet solution
11:08:17 <fsasaki> .. you trust your community and work with it
11:08:22 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
11:08:57 <fsasaki> mark: an example - using a volunteer translation
11:08:57 <fsasaki> .. you can take the wikipedia appraoch to take the latest translation
11:09:18 <fsasaki> .. my question about manage is: how do you choose the latest translation?
11:09:36 <fsasaki> reinhard: we have proofreaders  - we trust them more than editors
11:09:45 <fsasaki> gabor: every language communtiy in drupal has their own subsite
11:09:50 <fsasaki> .. the manage permissions
11:10:02 <fsasaki> .. we found out that they have very different mgmt styles
11:10:12 <RRSAgent> I'm logging. I don't understand 'draft  minute', fsasaki.  Try /msg RRSAgent help
11:10:14 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
11:10:43 <fsasaki> gabor: I found that in the open source world you need to have incentives to start work
11:10:49 <fsasaki> .. you need to have tasks that make people happy
11:11:00 <fsasaki> .. the micro contribution approach is helpful with that
11:11:04 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
11:11:33 <fsasaki> gabor: just read a book about incentives for people, how to make them happy doing that - very useful
11:11:42 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
11:12:35 <fsasaki> xyz: memorizing of music - at gavin: do you have experience with that?
11:12:53 <fsasaki> gavin: there are cantador, singers taht listen to each other while singing
11:13:05 <fsasaki> .. that active listeing while doing helps with memorizing
11:13:32 <fsasaki> xyz: students in germany have more and more problems to memorize what we did previously, so we need ways to help here
11:13:37 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
11:14:11 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
11:14:31 <fsasaki> chaals: many people talk about XLIFF - the web world uses json
11:14:43 <fsasaki> .. but tools use XLIFF etc. internally
11:14:54 <fsasaki> .. how do we manage the transition from XLIFF to json etc.
11:15:00 <fsasaki> .. how do we know when standards change?
11:15:14 <fsasaki> gavin: you have to be clear what is an declaration and what is an object
11:15:24 <fsasaki> .. a declaration is self-evident
11:15:30 <fsasaki> .. in json you can't declare
11:15:44 <fsasaki> gavin: we are using gettext, yaml, xliff, tools for ITS
11:15:50 <fsasaki> .. we use different tools for different problems
11:16:05 <fsasaki> .. we have a community trained and people using them
11:16:19 <fsasaki> .. if we don't see a beneift to move to new appraoches we won't move
11:16:25 <fsasaki> jan: agree
11:16:36 <fsasaki> .. we support many platforms, e.g. html5 windows web apps
11:16:45 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
11:17:04 <fsasaki> jan: how do we manage to export to something else than XLIFF
11:17:16 <fsasaki> .. when we look at services it comes artibtrary - can be xliff, json etc.
11:17:28 <fsasaki> .. so keeping the eye on the ball of standardization becomes crucial
11:17:38 <fsasaki> .. we have to continue whose conversations
11:17:44 <fsasaki> reinhard: XLIFF still has a long way to go
11:17:52 <fsasaki> .. sometimes standards can remove competition
11:18:02 <fsasaki> ... it is good to have competition in some areas
11:18:22 <fsasaki> .. XLIFF is trying to solve a problem that is not related to competition
11:18:33 <fsasaki> .. more uptake of XLIFF will help everybody
11:18:52 <fsasaki> .. it reminds me of character encoding discussion - now everybody is using unicode
11:19:06 <fsasaki> .. such developemnt allows to concentrate on more interesting problems
11:19:11 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
11:19:14 <fsasaki> lunch break
11:20:40 <laurent_oz> laurent_oz has left #mlwrome
12:34:23 <ivan> ivan has joined #mlwrome
12:43:59 <lmatteis_> lmatteis_ has joined #mlwrome
12:56:04 <tgraham> tgraham has joined #mlwrome
13:02:26 <fsasaki> fsasaki has joined #mlwrome
13:02:52 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki


13:03:54 <fsasaki> topic: Román Díez González and Pedro L. Díez-Orzas - ITS2.0 Implementation Experience in HTML5 with the "Spanish Tax Agency" Site
13:04:00 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:04:48 <fsasaki> roman: from spanish tax agency
13:04:51 <fsasaki> .. we are partners in the use case
13:04:54 <kfritsche> kfritsche has joined #mlwrome
13:05:44 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:06:59 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:06:59 <kfritsche> ??: why are we here - we are contributing with linguaserve to the MLW-LT project
13:07:11 <fsasaki> scribe: kfritsche
13:07:20 <fsasaki> s/??:/roman:
13:07:23 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:10:39 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:10:58 <Clara> Clara has joined #mlwrome
13:11:08 <GordonD> GordonD has joined #mlwrome
13:11:21 <fsasaki> scribe: fsasaki
roman: requires shifting to HTML5
...: automatic adding ITS2.0 data categorys by linguaserve
pedro showing demo
pedro: you can see here the global rules
...: added localization quality
pedro: we also have a poster outside for further information
pedro: the user doesn't see anything of the translation process, the translaiton happen in real-time
pedro: is aimed for big comapanies above 1mio words, otherwise its not profitable
roman: what the advantage of ITS2.0?
...: for us it is control
...: user has the control about the data cetegories
13:11:48 <fsasaki> pedro: client looses control over translation process in the tooling scenario
13:11:59 <fsasaki> .. the ITS2 metadata allows to re-gain control for the client
13:12:23 <fsasaki> roman: having as the client control  - that is what ITS2 is good for us
13:13:07 <fsasaki> .. we use various metadata items ("data categories"): translate, domain, localization note, locailzation quality issue, mt confidence, provenance
13:13:12 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:14:04 <fsasaki> pedro: page in various languages
13:14:18 <fsasaki> .. ITS2 metadata specifies what can be translated or not
13:14:31 <fsasaki> .. we have also localization qualtiy issue created by the post editor
13:14:56 <fsasaki> .. we also need to develop related best practices about using ITS2 e.g. by post editors
13:14:56 <kfritsche> kfritsche has joined #mlwrome
13:15:04 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:15:44 <fsasaki> roman: shfiting to HTML5 various steps
13:15:57 <fsasaki> .. shallow HTML5: oboslete attributes
13:16:02 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:16:22 <fsasaki> .. second, automatic annotation
13:16:28 <fsasaki> .. third, manual annotation facilities
13:17:16 <fsasaki> .. example of domain name annotation - tagging was done by scripts
13:17:29 <fsasaki> .. provide an editor for manual annotation
13:17:36 <fsasaki> pedro: last 30 seconds:
13:17:43 <fsasaki> .. next steps is to end the use case
13:17:51 <fsasaki> .. it is complete and functional
13:18:16 <fsasaki> .. exploring best practices is a critical topic, once the standard its2 itself is finished
13:18:42 <fsasaki> .. there are other metadata items like "readyness" that are not part of Its2 but which can be extensions to its2
13:18:50 <fsasaki> .. then methodologies for post editing
13:19:03 <fsasaki> .. and specific tools for dealing with the metadata are needed


13:19:30 <fsasaki> topic: Hans-Ulrich von Freyberg - Standardization for the Multilingual Web: A Driver of Business Opportunities
13:19:35 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:20:15 <fsasaki> hans: MLW-LT is not only for geeks, but also for accountans
13:20:52 <Clara> Clara has joined #mlwrome
13:21:01 <fsasaki> .. about cocomore: communication and IT, largest drupal dev teams in germany and spain
13:21:06 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:21:47 <fsasaki> hans: why do we engage in MLW-LT? we want to lay the foundation of aa real integration of a CMS in the localization chain
13:21:57 <fsasaki> .. our role in the MLW-LT project: contributing to the ITS2 standard
... we had to main task in the ITS
... first enchance drupal to use ITS2.0
... second to create a rountrip for translations
13:22:10 <fsasaki> .. and enhancing drupal to work with ITS rules
13:22:43 <fsasaki> .. also creating a use case to demonstrate that ITS2 creates business benefits
13:22:52 <fsasaki> .. not only for localization but business at large
13:23:17 <fsasaki> .. example client vdma - industrial assocation in Germany of exporting companies
13:23:29 <fsasaki> .. export business means multilingual challenge
13:23:42 <fsasaki> .. today vdma has to handle 9 European languages
... VDMA one of our customers, want to translation into many languages
... translation needs time and has to be managed
... in our use-case ITS2.0 is used
13:23:53 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:24:15 <fsasaki> hans: vdma has to publish online and offline for 60 sub sectors = domains that have to be covered
13:24:27 <fsasaki> .. they have a central product database that also has to be multilingual
13:24:43 <fsasaki> .. all has to be managed
13:25:09 <fsasaki> .. we said to vdma that ITS2 / mlw-lt metadata can help the vdma business
13:25:32 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:25:44 <fsasaki> hans: implementations in drupal for ITS2:
13:25:49 <fsasaki> .. its rules to be used in drupal
... we have a drupal module to add local and global ITS data category
... developed a language management system to edit ITS data without touching content
... extended tmgmt to handle ITS and can export strings
... showing data categories in WYSIWYG editor
13:26:01 <fsasaki> .. a wysiwyg edior for applying ITS rules
13:26:13 <fsasaki> .. and we implemented a transation mgmt tool
13:26:24 <fsasaki> .. it allows to view and edit the metadata without a cms system
13:27:36 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:27:50 <fsasaki> hans: now more about the use case - annotation editor screen shot
13:28:12 <fsasaki> .. it allows to deal with items that have been defined in ITS 2.0
13:28:36 <fsasaki> hans: export of XML file to linguaserve
... we are allowing domains
... exporting to linguaserve
13:28:49 <fsasaki> .. here how the XML arrives at linguaserve
13:29:06 <fsasaki> .. in the format you have XML ITS 2.0 metadata
13:29:26 <fsasaki> .. after translation the whole information returns to Cocomore, in the CMS for review
13:29:29 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:29:46 <fsasaki> hans: two more interfaces in the CMS
13:29:54 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:30:18 <fsasaki> hans: translation process overview
13:30:23 <fsasaki> .. interface of language mgmt tool
13:30:40 <fsasaki> .. tool to view and edit the metadata, the jquery plugin
... reintegrated content can be reviewed
... translation process overview to see the status of the translation foreach element
... time is money
... difference between manual and automatic processing of a translation
13:32:06 <fsasaki> hans: options of saving time
13:32:19 <fsasaki> .. translation process steps that have to be done from the client point of view
13:33:21 <fsasaki> .. from the client and LSP point of view we analysed the process of an LSP
13:33:24 <fsasaki> .. e.g. receiving, storing data etc.
13:33:39 <fsasaki> .. processing annotation information etc.
13:34:03 <fsasaki> .. in all process steps a lot of time saving
... 84% of time is saved
... we also have this for the LSP point of view
... 83% of time is saved
... standardization saves money
... reduced the overhead and not the translation time
13:34:26 <fsasaki> .. conclusion: standardization combined with automatic annotation and round tripping was 80%
13:34:41 <fsasaki> .. we also reduced the time line
13:34:50 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:35:09 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:35:22 <kfritsche> kfritsche has joined #mlwrome


13:35:41 <fsasaki> topic: Brian Teeman - joomla: Building Multilingual Web Sites with Joomla! the Leading Open Source CMS
13:35:45 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:36:37 <fsasaki> s/bryan/brian/
13:37:24 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:42:56 <fsasaki> ok, vielen Dank
13:43:00 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:43:56 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:45:52 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:47:01 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:47:55 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:51:37 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
13:54:54 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
14:13:45 <omstefanov> omstefanov has joined #mlwrome
14:22:04 <leroy> leroy has joined #mlwrome
15:10:15 <fsasaki> fsasaki has joined #mlwrome
Brian:language matters
... in 65+ translation available
... some created by teams, some by single persons
... en-gb is the default, but not required
... mono-lingual works
... multi-lingual, also no problem
... tag, translate and associate content
... joomla 3 mobile ready, one site all sizes
... you can checkout multilingual joomal sites juv.nu and hotelreoyalsavoymadeira.com
... joomla is easy to extend
... community translation support to translate your own customiziation
... developer can use opentranslator.org
... enhance your site with over 6000 extras
... learnjoomla3.com is english only
... ayudajoomla3.com would be better for spanish content
... example 2 - translation managment system (module called josetta)
... overview of translated content
... see translation and original side by side
... here for questions for both days



Topic: Vivien Petras - The Europeana Use Case-Multilingual and Semantic Interoperability in Cultural Heritage Information Systems
vivien:we want every thing from cultural haritage
... many different languages, english issn't even under the top 5
... users use their own language
... many collections from germany, therefor many meta tags in german
... even if there are many pictures most of the people only look at items, with meta tags with ther native language
... search is not multilingual aware, same results for all languges
... music and images has no language, therfore most objects are in the "Multilingual" language
... because most user only look at native langugages, most of the items maybe never get clicked
... user want to search for their languages
... we could use semantical enrychment to improve multilingual search 
... but this led to other problems
... new enrichtment plan to link to contextual vocabularies from providers
... Europeana is now Open Data, we provide a sparql endpoint and RDF download



topic: Inna Nickel, Christian Lieske and Daniel Naber - tool suported linguistic quality in web-related multilinguage context
christian: linquistic quality is highly scenario dependant and never the same
... requirements are very different in style, voice and terminology
... integration into openoffice
... sample the single word Link - what is the meaning of this?
... NLP is about voice control and machine transaltion and linguistic quality check
... tooling to support high quality translation is
Inna: open source tool for SLQ - LanguageTool supports ITS
... LanguageTool can be used standalone, embedded in java or could be use through OKAPI over HTTP
... firefox plugin to do quality report directly in the browser
... implemented russia data rule set from enterprise scenario
... limitation is processing of languagetool and complexity of enterprise data
... language tool has also the ability to check if a homepage is simple to read, which is imprortant to have better accessibility


Q/A:Creators

???: is joomla support language tools
brian: no, does drupal?
gabor: we building a framework, which would make it easy to use any
felix: jquery plugin would be CMS independent

felix: when will joomla will support ITS or XLIFF
brain: when someone will implement it - we have no company behind it - when there is a use for that
hans: in joomla its most private and small companies, drupal have high level entrprise user, which could have more usage of ITS

brain: interferences between jQuery Plugin between TMGMT
hans: no problems between jQuery Plugin and TMGMT, as both addresses different fields




15:10:39 <fsasaki> topic: bryan schnabel - Making the Multilingual Web Work: When Open Standards Meet CMS
15:10:47 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:14:08 <fsasaki> scribe: fsasaki
15:14:12 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki



15:15:10 <fsasaki> bryan: multilingual web needs multilingual content
15:15:20 <fsasaki> .. I focus on drupal CMS and XLIFF extensions to that
15:15:55 <fsasaki> .. drupal has an out of the box solution - translation core module
15:16:05 <fsasaki> .. "translator logs into drupal"
15:16:14 <fsasaki> .. they click "translate" tab
15:16:21 <fsasaki> .. choose the language of translation
15:16:33 <fsasaki> .. add a new translation etc.
15:17:34 <fsasaki> .. a unique name .. boomb!
15:17:44 <fsasaki> .. pros of out of the box solution: easy to use
15:17:59 <fsasaki> .. cons: translator can't leverage translation memory
15:18:10 <fsasaki> .. needs access to drupal cms
15:18:13 <fsasaki> .. and he can do harm
15:18:23 <fsasaki> .. so idea is to leverage drupal with XLIFF
15:18:33 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:18:58 <fsasaki> bryan: in the xliff scenario, the roles are like this:
15:19:07 <fsasaki> .. drupal admin selects the node types to translate
15:19:12 <fsasaki> .. he exports to xliff
15:19:17 <fsasaki> .. saves it to harddrive
15:19:36 <fsasaki> .. some times post processing is needed
15:20:05 <fsasaki> .. to avoid change of text <target state="final"> is set
15:20:14 <fsasaki> .. XLIFF is then sent to LSP
15:20:18 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:20:29 <fsasaki> .. so the translator doesn't get drupal nodes, but XLIFF
15:20:39 <fsasaki> .. translator can use TM, their translation tool etc.
15:20:57 <fsasaki> .. I got what I wanted:
15:21:06 <fsasaki> .. translation is like I wanted
15:21:23 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:21:38 <fsasaki> bryan: some strings are not part of the xliff model
15:21:41 <fsasaki> .. e.g. UI strings
15:22:17 <fsasaki> .. the drupal admin can export these as PO files
15:22:36 <fsasaki> .. and there is a workflow PO > XLIFF , translation > PO , then into drupal again
15:22:42 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:22:53 <fsasaki> bryan: advantage of approach:
15:23:08 <fsasaki> .. better for translator
15:23:19 <fsasaki> .. now about CCMS Component content management system
15:23:32 <fsasaki> .. CCMS uses topic based authoring
15:23:42 <fsasaki> ... DITA lets you have topics, maps, images etc.
15:23:47 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:24:05 <fsasaki> bryan: we could talk about thousands or millions of files
15:24:53 <fsasaki> .. now workflow in CCMS
15:25:18 <fsasaki> .. the DITA aware nows about the translation related properties of content
15:25:49 <fsasaki> .. coming back to the two scenarios: in the first scenario, we put everything into a ZIP file and the LSP has to deal with this
15:26:06 <fsasaki> .. the disadvantage is that they have to deal with millions of topics to translate
15:26:11 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:26:43 <fsasaki> bryan: now augmenting the process by using XLIFF
15:26:51 <fsasaki> .. see the XLIFF dita open toolkit plugin
15:27:20 <fsasaki> .. integration with trisoft / visual studio
15:27:44 <fsasaki> .. now with new workflow: 1 xliff file, not millions of DITA files sent to LSP
15:27:56 <fsasaki> .. advandage: LSP don't have to know dita
15:28:09 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:28:18 <fsasaki> bryan: disadvantage: it is complex
15:28:40 <fsasaki> s/complex/complex (scribe missed)

15:28:58 <fsasaki> topic: sinclair morgan - How do you publish one thousand web pages, in 12 languages, at a high quality, 50% quicker than you can today?
15:29:01 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:30:52 <fsasaki> sinclair: want to emphasize how machine translation can help to manage high volume of content in a high quality translation process
15:31:20 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:33:05 <fsasaki> sinclair: three scenarios - conventional translation from scratch, with TM, and with MT + post editing
15:33:33 <fsasaki> .. scenario 1): 2500 words per day per translator. Total costs = 86.000 Euro
15:33:47 <fsasaki> .. scenario 2): 42.500 Euro
15:34:13 <fsasaki> .. assuming 50% leverage
15:34:29 <leroy> leroy has joined #mlwrome
15:34:37 <fsasaki> .. scenario 3): trained MT + PE: many variables are important
15:35:03 <fsasaki> .. e.g. MT output quality, language direction, quality of the source content, the training content, translation environment etc.
15:35:16 <fsasaki> .. snapshot of recent term projects:
15:35:33 <fsasaki> ... very dependent on projects and content
15:35:49 <fsasaki> .. average improvemen tof 43%
15:36:02 <fsasaki> .. so that's 30.000 Euro
15:36:24 <fsasaki> s/improvemen tof/improvement of/
15:36:32 <fsasaki> .. in scenario 3) we use both MT and TM
15:36:42 <fsasaki> .. we have a team of post editors, not translators
15:37:03 <fsasaki> .. so we had in 1) 40 days, in 2) 30 days, in 3) 14 days
15:37:09 <fsasaki> .. and the above cost savings
15:37:14 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:37:33 <fsasaki> sinclair: what is needed for this performance?
15:37:40 <fsasaki> .. you need MT system with high quality
15:37:46 <fsasaki> .. a set of baseline languages
15:37:56 <fsasaki> .. you need a large amount of data to build trained enginges
15:38:06 <fsasaki> .. you need a system that is easy to minister and secure
15:38:15 <fsasaki> .. MT adaptation is important for specific applications
15:38:31 <fsasaki> .. you need to be able to integrate MT system into your environment
15:38:38 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:39:08 <fsasaki> sinclair: comparison between convential translation versus baseline vs. trained mt systems
15:39:19 <fsasaki> .. the trained mt system scores much better than a baseline system
15:39:25 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:40:10 <fsasaki> sinclair: productivity increase is important
15:40:30 <fsasaki> .. integration with translation environment
15:41:38 <fsasaki> .. easy to use interface, TM and automated workflow, support functions (terminology, spell checker, qa tools, reporting)
15:42:02 <fsasaki> .. efficient integration means: retain all benefits that we have in an MT environemnt, and add the savings from MT
15:42:06 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:43:09 <fsasaki> bryan: example of baseline + trained MT engine - quality of trained MT engine is clearly better
15:43:28 <fsasaki> .. human resource: need MT developers, MIT linguists, post-editors
15:43:38 <fsasaki> .. post-editing is a professional skill
15:43:47 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:44:14 <fsasaki> sinclair: how to guarantee high quality: use the same qa checks as for human translation
15:44:27 <fsasaki> .. include a linguistic review of the post-editing work
15:44:40 <fsasaki> .. use the same qa standrd as conventional translations
15:44:53 <fsasaki> .. review gains of process
15:45:03 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:45:47 <fsasaki> need a language technology infrastructure platform
15:45:55 <fsasaki> .. productiiy, quliaty, automat worfklw
15:46:08 <fsasaki> .. SMT, baselines, fast and csot efficient rainign
15:46:12 <fsasaki> .. easy to deploy MT
15:46:24 <fsasaki> .. and underpinnoing that with human resources
15:46:43 <fsasaki> s/need/sinclair: summary: need
15:46:47 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:47:02 <fsasaki> s/productiiy/productivity/
15:47:10 <fsasaki> s/quliaty/quality/
15:47:23 <fsasaki> s/automat worfklw/automated workflow/
15:47:43 <fsasaki> s/csot efficient rainign/cost efficient training/
15:47:53 <fsasaki> s/underpinnoing/underpinning/
15:47:57 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:48:54 <fsasaki> s/productivity/including - productivity/
15:48:55 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki

15:49:13 <fsasaki> topic: Charles McCathie Nevile  - Localization in a Big Company: The Hard Bits
15:49:17 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:51:01 <fsasaki> chaals: introducing yandex - major russian search engine
15:51:10 <fsasaki> .. providing all kinds of services running in russian
15:51:18 <fsasaki> .. companies DNA comes from language technology tools
15:51:28 <fsasaki> .. the company was started based on russian morphology analysis
15:51:37 <fsasaki> .. they applied these to search
15:51:49 <fsasaki> .. originially this was just for russian
15:51:53 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:52:52 <fsasaki> chaals: our search results page has 1-2 results for each page, for a particular user in a given time + place
15:53:20 <fsasaki> .. we try and give you local results, but we try not to do too much of that
15:53:26 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:53:41 <fsasaki> chaals: we try not to personalize too much, people don't like that
15:54:58 <fsasaki> .. linguistic processing of russian: gender issues, case issues
15:55:03 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:56:28 <fsasaki> chaals: some services are localized, focus still on russia
15:56:41 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
15:56:52 <fsasaki> example of yandex homepage
15:58:31 <fsasaki> home pages for Russia and Kazakhstan
15:58:47 <fsasaki> chaals: BEM (block-element-modifier)
15:58:59 <fsasaki> .. open source library to let front end developers to put together a new page
15:59:31 <fsasaki> .. yandex uses it for users settings, e.g. flags per language
15:59:44 <fsasaki> .. we are using flags and iso language codes
15:59:57 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:00:32 <fsasaki> chaals: different content in different languages, different design asthetics
16:00:41 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:00:52 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:01:49 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:02:13 <fsasaki> chaals: every company has an internal bias
16:02:26 <fsasaki> .. language technology is yandex DNA
16:02:51 <fsasaki> .. we are building on top of that has been developed so far
16:02:59 <fsasaki> .. standards are important, we recently joined w3c
16:03:18 <fsasaki> .. our developers speak and read and write .. in russian
16:03:26 <fsasaki> .. that is something the company has to deal with
16:03:33 <fsasaki> .. thank you for your attention
16:03:51 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki

16:07:38 <fsasaki> topic: hans uszkoreit - Quality Translation: Addressing the Next Barrier to Multilingual Communication on the Internet
16:08:06 <fsasaki> had to pass Manish Kanwal presentation, due to connectivity problems
16:08:13 <fsasaki> s/pass/pass by/
16:08:24 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:08:45 <fsasaki> hans: other speakers are from business and talking about success story
16:09:06 <fsasaki> .. in business if you are dissatisvied too long you are out of business
16:09:16 <fsasaki> .. in research if you are satisfied too long you are out of business
16:09:40 <fsasaki> .. MT success stories: free only MT systems, in-house online MT systems
16:11:20 <fsasaki_> fsasaki_ has joined #mlwrome
16:11:32 <fsasaki_> scribe: fsasaki_
16:11:42 <fsasaki_> hans: there are MT sucess stories
16:11:47 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki_
16:12:17 <fsasaki_> .. closely related languages work well, but others and speciaized languages don't work
16:12:31 <fsasaki_> .. MT translation research has concentrated on high volumes
16:12:38 <fsasaki_> .. needed for inbound translation etc.
16:12:41 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki_
16:13:20 <fsasaki_> hans: there is a lack of translation quality for outbound translation
16:13:27 <fsasaki_> .. let's take a new appraoch:
16:13:45 <fsasaki_> .. separation in good enough, almost good enough and not usable for outbound
16:13:48 <fsasaki_> scribe: fsasaki
16:14:15 <fsasaki> hans: good can be 5-75%, then 15-65, then 5-75
16:14:20 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:14:57 <fsasaki> .. increases in bleu score: currently it is gained in the "red" part, that is "not usable"
16:15:19 <fsasaki> .. the new current approach tries to recognize truly good high quality estimation
16:15:24 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:15:53 <fsasaki> .. in the middle is compter assistant translation - how to move the "almost good" into the "good" area
16:16:09 <fsasaki> .. many projcets now help to move to the left, e.g. wrt to post-editing
16:16:12 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:16:49 <fsasaki> hans: we tried to push the topic with various instruments
16:17:19 <fsasaki> .. one is META-NET a network of excellence that created a vision process in line with the preparation of the EU horizion 2020 program
16:17:54 <fsasaki> .. we focused on 31 languages in Europe
16:18:18 <fsasaki> .. see http://www.meta-net.eu/whitepapers
16:18:42 <fsasaki> .. then there is an infrastructure for sharing resources, see http://www.meta-net.eu/meta-share/
16:19:03 <fsasaki> .. and there is a strategic research agenda, see http://www.meta-net.eu/sra
16:19:42 <fsasaki> .. the SRA describes the needs of the industry, the predictions, mega-trends, research priorties
16:19:47 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:19:57 <fsasaki> hans: strategic considerations:
16:20:28 <fsasaki> .. we sid we woudl concentrate on some areas which have a high chance of being succesful for Europe
16:20:37 <fsasaki> .. three research topics for the SRA
16:20:58 <fsasaki> .. translingual cloud, social intelligence, socially aware interactive assistants
16:21:06 <fsasaki> .. all topics are highly interconnected
16:21:11 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:21:46 <fsasaki> hans: a lot of technology is used by several groups as the same time, e.g. a text parser
16:22:01 <fsasaki> .. about the translingual cloud: can be a method of generic and special-purpose checking
16:22:16 <fsasaki> .. automatic translation, language checking, post-editing etc.
16:22:29 <fsasaki> .. systematic concentration on quality barriers
16:23:04 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:23:24 <fsasaki> .. ingredients: semantic translation paradigm
16:23:45 <fsasaki> .. exploitation of strong monolingual natural language analysis and generation
16:24:06 <fsasaki> .. and modular combination of specialized analysis
16:24:45 <fsasaki> .. european service platform: in addition to the strategic research agenda, there is a proposal of a platform for services not restricted to translation
16:24:49 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:25:09 <fsasaki> .. you can hook into services that are not yet multilingual
16:25:25 <fsasaki> .. finally about a project that is a pilot to prepare something: QTLaunchPad
16:25:42 <fsasaki> .. assemble data and tools
16:25:49 <fsasaki> .. create a shared quality metrics
16:26:03 <fsasaki> .. has been demoed and is now being finished
16:26:15 <fsasaki> .. then extending existing platforms for sharing
16:26:51 <fsasaki> .. consortium comprises DFKI, CNGL DCU, ILSP athena and Univ. of Sheffield
16:26:56 <fsasaki> .. as a subcontractor GALA
16:27:02 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:27:09 <fsasaki> .. QTLP planning panel:
16:27:21 <fsasaki> .. has many names in the European translation industry
16:27:31 <fsasaki> .. important bit is semantics based translation
16:27:48 <fsasaki> .. already in 1949 that was a vision by Warren Weaver
16:28:14 <fsasaki> .. now in statistical MT, it is clear that we have to go into semantics deeper
16:28:23 <fsasaki> .. this can be a talk of people working in semantic web
16:28:33 <fsasaki> .. if you look into the work on semantic web
16:28:43 <fsasaki> .. there is a stuff that can be used for semantically used MT

16:29:10 <fsasaki> topic: localizers Q/A
16:29:15 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:31:03 <fsasaki> questions from Paula Shannon
16:31:08 <fsasaki> scribe missed
16:31:21 <fsasaki> sinclair: cost calculation is still an issue
16:31:34 <fsasaki> .. if MT is provided as a service there is no additional cost
16:31:48 <fsasaki> hans: my answer to my question
16:32:01 <fsasaki> .. in a spirit what I proposed is close to "knowledge based translation"
16:32:16 <fsasaki> .. but it is diffierent: the approach to semantics has changed completely
16:32:39 <fsasaki> .. in the past the idea was that people sit together and build enhanced semantic modles
16:32:54 <fsasaki> .. but now we are getting to applications based on also crowed sourced applications
16:33:13 <fsasaki> .. now people find huge collections that can be used as interlingua
16:33:23 <fsasaki> .. so it will come, but in a totally different dress
16:33:33 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:34:30 <fsasaki> various questions from tomas
16:35:57 <fsasaki> bryan: the zip file approach - we never did it that way
16:36:45 <fsasaki> dan: specialzed looking for language universals found only 6-7 universal
16:36:52 <fsasaki> .. this is not enough for an MT system
16:37:00 <fsasaki> hans: a very good piont
16:37:08 <fsasaki> .. highly specialized systems in speeach and MT
16:37:18 <fsasaki> .. the systems do not perform better than the generic one
16:37:29 <fsasaki> .. but that doesn' mean that this has to stay
16:37:37 <fsasaki> .. it only shows that we are doing things wrong
16:37:42 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:37:57 <fsasaki> hans: this shows just that we don't get various models right
16:38:12 <fsasaki> .. if you talk to developers of big systems
16:38:23 <fsasaki> .. right now, because of huge amount of data
16:38:34 <fsasaki> .. the generic system is still ahead
16:38:49 <fsasaki> .. but I would bet quite a bit of money that the specialized systems will be better
16:38:53 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:39:05 <fsasaki> hans: that is why we leave the space for many companies
16:39:11 <fsasaki> .. that google and bing will never enter
16:39:21 <fsasaki> chaals: not so sure
16:39:44 <fsasaki> .. yandex has no reason to go away from any area of translation
16:40:01 <fsasaki> .. and not to augment our tools with tweaks that will make an improvement
16:40:07 <RRSAgent> I'm logging. I don't understand 'draft minute', fsasaki.  Try /msg RRSAgent help
16:40:09 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:40:32 <fsasaki> hans: this is not a technical question
16:40:42 <fsasaki> .. of course a generic system can do what the specialized system does
16:41:08 <fsasaki> .. but it is a question of whether the big players then it comes to semantics will have all the knowledge of every subfield
16:41:15 <fsasaki> .. and if they will share it
16:41:56 <fsasaki> .. if it is possible in this more semantic age to have more modeling and that are not shared, then there is a possibiltiy that the business of the smaller companiesis succesful
16:42:04 <fsasaki> s/companiesis/companies is/
16:42:28 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:43:18 <fsasaki> gavin: missed question
16:43:35 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:45:37 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki
16:47:11 <fsasaki> wrapping up day one
16:47:15 <RRSAgent> I have made the request to generate http://www.w3.org/2013/03/12-mlwrome-minutes.html fsasaki