<stevelee> hi roy - webEx is getting me to start the meeting - did somehting change?
<JF> https://github.com/w3c/personalization-semantics/wiki/July-8th-notes-on-Options-for-symbol
JF: concern for Bliss -Creative
Commons
... symbol own taxonomy with a unique ID
... bliss symbols spreadsheet, ~6K symbols
... ref #, term, definition of term
... conflate crocodile / alligator for example
... intending to do, use their reference # as the value of data
symbol.
... think of it as key # first we will have a subset of those
6K terms.
... analogy - css hex value you can also shorter list of name
of a color
... FF0000 or red for example
... we will reference # from Bliss but other symbol sets
against the Bliss #
... Bliss seems to be the largest set
... borrow their list and we use their key # for us as
well.
Steve: built up sentences from these bits of symbols?
JF: not a ton of hand authored
content, there will be tools (cart transcription keyboard) key
strokes that building it up (stenographer)
... converting text to symbols it wasn't a replacement, the
symbols were written on top of text.
... Chaals suggested this sounds like "ruby" annotations.
... expands the line spacing to insert the symbols. not our job
about the presentation layer
Steve: master set of words, it
looks like its probably best we have.
... nouns are easy but verbs are harder etc.
Janinia: this is the best in class (Bliss)
JF: we don't pick winners, but it
appears to be one of the largest sets and still active and
updating, and robust to add new symbols.
... concern on licensing.
Janina: funding behind connection funding translation between symbol sets (Some UN / EU agency.
JF: between now and TPAC / with
6K symbols the bliss taxonomy is robust we choosing to use
their taxonomy or that is the proposal.
... master # bliss # see the # and definition, 12322 or you can
use a common term. (4 possible terms into 1 symbol)
... they have a number they have definition and a symbol.
Janina: I think there is agreement
Steve: it does say this is a noun, there may be another definition for verbs as well
JF: just wanted to see if this
"Can" work.
... , 12345 def (advice, council, recommendation) shorter list
of human readable terms is what lisa was thinking.
Janina: where they ran into issues from Matt is noun from verb. we can tell them that they need to fix it to go mainstream.
JF: there was nouns and verbs, as well as ajectives from the Word Net reference.
Janina: if their definitions are inaccurate, we need it to work and not confusing people. Close isn't good enough. Significant # of people this must be really right.
<stevelee> Here are the mulberry symbols - purely for reference as an open symbol set (cc by)
<stevelee> https://mulberrysymbols.org/assets/categories/categories.pdf
Janina: yes Bliss is a useful set, but can we use it to do something useful.
JF: we are not expecting machines
to do the translation initially. initially common terms list
will be easier for humans to edit.
... , not to create translation, point to provide a means to
add the metadata, but initially would be human.
Janina: only some authors can do this this requires expertises.
JF: no argument on the authoring
level this will require expert human understanding to use
this.
... timestamp around each word, all we are doing and when your
machine does that and there is the data-symbol and then the
expert will look up the correct reference #.
... , red vs. I read a book. (noun vs. verb) I will assume
Bliss has solved or in the process to solve this. red 12345,
and read verb is 456789, and the author must decide which ref #
to use.
... , not our job to know what symbol to use, we are building
tools not content. its the job of the author to pick the right
#
... , Bliss, 1. largest, 2. International in scope, 3. active
dev., 4. liberal license
... , we need to pick a symbol set
Steve: similar to Unicode,
unicode was decided it wasn't feature complete, so changes are
inevitable.
... , AACU users you can support them, then people can work
with that.
JF: other concern / problem there
are multiple symbol sets we can infer one, ie. bliss, or
unicode .. how do we tell the user to tell what symbol set we
are using? how do we link our symbol table to the document.
Lisa suggested linkrel to personalization and here is the
data.
... , using schema.org accessibility feature accessibility term
(symbolset) but schema.org could point to a full URL. when we
want to use a #, definition , symbol. we get out of the way of
verbs / nouns
Janina: what are we building?
JF: we are providing a markup mechanism to select a word/phrase, concept) is referenced by a specific symbol by using a specific ref#
Janina: could be a proxy server / hand authored.
JF: this ref # i will go to my
symbol set and provide the symbol to represent that
... , we are sold with A) how do we reference the symbol set
(ie. Font family idea)
Charles: we need to still reference which symbol set we are using.
JF: we didn't want to use micro
formats for purpose of inputs. we did not have
taxonomies.
... we borrowed one persons taxonomy ie autocomplete already,
similar to Bliss
... we need a taxonomy min: ref id, definition term, and the
symbol.
... if we just allow each symbol set.
... , we want to make sure that if we tie ourselves to bliss
but if Bliss decides to do something different we need to make
sure pick wordnet/bliss/ or some other we need to make the
right smart choice.
... , we need a Key # we need to have a powerful bases in order
to get it right.
... , are there any legal reasons not to use Bliss.
Janina: we need a better choice?
JF: who is going to do this reasearch.
Charles: we can ask Matt back (maybe Abby suggested by Steve)
JF: Lisa wanted to use a common term for some of the concepts instead of a number, red/vs ff0000
Janina: then parsing is a lot trickier.
JF: one of the requirements while # is more adjustable, name values is weaker, Lisa requested easy of authoring that is important. how many do we want to reference using a human writable
Steve: will some authors would they want to use a name vs. #
JF: who ones the mapping table
and how many.
... , we need to talk to the internalization folks (Richard
Yachita) at TPAC
... , Check to see if we all agree that schema.org metadata
might be the best approach to link to the specific dataset
being use which can be overridden by the User Agent to map to a
different symbol set.
Charles: need to add "symbols" to the accessibilityFeature list of values, and then a new schema.org attribute with a reference to the symbol set used (which can be overridden by the end user if desired) Discuss this at our next meeting.
<scribe> ScribeNick: CharlesL
<scribe> Meeting: Personalization Task Force Weekly Meeting
This is scribe.perl Revision: 1.154 of Date: 2018/09/25 16:35:56 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00) Present: CharlesL janina JF Roy stevelee Regrets: Lisa Becky Found ScribeNick: CharlesL Inferring Scribes: CharlesL WARNING: No "Topic:" lines found. Found Date: 15 Jul 2019 People with action items: WARNING: Input appears to use implicit continuation lines. You may need the "-implicitContinuations" option. WARNING: No "Topic: ..." lines found! Resulting HTML may have an empty (invalid) <ol>...</ol>. Explanation: "Topic: ..." lines are used to indicate the start of new discussion topics or agenda items, such as: <dbooth> Topic: Review of Amy's report WARNING: IRC log location not specified! (You can ignore this warning if you do not want the generated minutes to contain a link to the original IRC log.)[End of scribe.perl diagnostic output]