Minutes: MathML intent meeting 17 Feb, 2022

 Attendees:

   - David Carlisle
   - Sam Dooley
   - David Farmer
   - Deyan Ginev
   - Moritz Schubotz
   - Murray Sargent
   - Neil Soiffer
   - Bert Bos
   - Paul Libbrecht
   - Bruce Miller
   - Patrick Ion
   - Cary Supalo

Regrets:

   - Steve Noble

Announcements/updates

BB: TPAC 2022 is tentatively scheduled for 12-16 September in Vancouver
Canada as a hybrid meeting accommodating both in-person and remote
attendance. This survey
<https://www.w3.org/2002/09/wbs/34786/tpac2022-fall-meetings/> will help
W3C in its decision to go ahead with this plan.

BB: This questionnaire is open for answers until 23:59, Boston time on
2022-03-28. Answers received after that time may not be counted.

*ACTION:* NS will fill out this form.

WE had a discussion on who would like to go in person, and who would like
to go virtually.

NS: Asked people who would go virtually to commit to four to six hours a
day for three days to the TPAC meetings.

Six people thought they would go in person. Four people thought they would
go virtually.
Intent: next steps?

DG's list of intents can be seen at DG-List Of Intents
<https://docs.google.com/spreadsheets/d/1EsWou1K5nxBdLPvQapdoA9h-s8lg_qjn8fJH64g9izQ/edit#gid=1358098730>
.

PL: For us to see more through the list of intents and be sort-agnostic, I
have created a little “display” of the intents that are gathered inside the
table of level 1 with a little heuristic trying to show a “big symbol” for
each intent. It can be seen here: (
http://notations.hoplahup.net/IntentsList/index.html). The code is based on
a csv export of the table followed by the execution of this script. All
this will move to a repo soon. It does not, yet, display the whole
information contained in the table.

PL: The link labeled "K" takes you to the Khan Academy reference.

PL: In the table, "G1 math" stands for mathematics used in the first grade
in the U.S. Children are generally six years old when they come into this
grade level.

NS: Do we have to distinguish between the G1, G2, and G3 levels? Cannot we
call them all by one name?

DC: These numbers may mean nothing to people from other countries.

NS: US School grade levels may not be a good way to specify different
subjects.

SD: Different states have different standards for different grade levels.

PL: Perhaps we could use student ages.

NS: Why do we care if it is G2 or G3? NS would like to use more broadly
defined subject areas.

PI: "G" can mean group and not grade.

DC: What does G7 mean in terms of age?

NS: It means age 12. Add five years to the "G" number.

DG: It is a mess to organize this. Perhaps we should not be so granular.

NS: Perhaps we should use the subjects: "beginning math", Algebra,
Geometry, and Precalculus.

NS: Would like to have around five categories.

NS: He is not clear what goes into pre-calculus.

SD: If it is not algebra or geometry, and is needed for calculus, it is
precalculus.

NS: Likes broad categories.

NS: How do you benefit from categories? Where does the subject area help
disambiguate things?

NS: He might add probability and statistics. He would Use subject area for
disambiguation.

NS: In the U.S. "Kilometers" are in an advanced subject area :-)

PL: He could add search to his display.

DG: In the case of "feet:, There is a difference between who owns a unit
(G1 Math) and where the unit is used ("used in the U.S).

DG: Suggested reducing the subject area differences.

BM: What do we want to achieve with organizing and categorizing things?

PL: We want to make shorter lists that a human can see easily.

NS If we assign a subject area that has not been annotated, would we be
able to guess the correct notation that would be used for the subject?

NS: Wants to use the subject area to help with notation.

BM: Use machine learning to look through a list of all possible notations
to figure out which are the usefully distinct subject areas. For example,
G4 and G5 are not distinct subjects.

BM: Parentheses have different meanings in different subject areas.

NS: DF has access to open-source textbooks. He sent ten textbooks to NS. NS
used these books to decide what were the most frequently used symbols. He
used the info to put those symbols in a smaller table so MathCAT starts up
quickly. If the smaller table does not have the symbol, the MathCAT looks
in much larger full table of symbols.

NS: distinguishes between symbols like the greater than symbol, and symbols
like "T" which, as a superscript, could stand for the transpose operation.
He considers them "notation"

PI: Distinguishes between more complex symbols like polar coordinates, and
superscripts like "t" (transpose).

NS: wondered if computers could distinguish symbols from notations in DG's
tables.

DG: thought that this would be difficult.

DG: His table has a column called "form". You could do searches on that
column.

DG: Look at real-world examples to see what is used, and if we need to have
a subject to cover it.

NS Categorizing things will help with making intent parameters.

NS: What should we do next?

DG: Perhaps we should define subject areas in different countries.

NS wants the group to suggest things that should be added or deleted from
DG's tables.

PL: It is too early because we may not have completeness.

NS: to push the project along, NS is working on MathCAT.

PL: We could assign people subjects to see if things should be added or
deleted to DG's tables.

PL: A person could take a book and see if we have something missing from
DG's sheet.

DG: If you give DG a list of symbols, he can compare that list to what is
in his tables rapidly.

DF: He has seventy-one open-source books in his list
<https://pretextbook.org/catalog.html>.

*ACTION:* DG will look through DF's books to see if DG should add symbols
to DG's tables.

DG: It is hard to decide what is in level one versus higher levels.

DF: Most of the books are for grades K through fourteen.

DF: I don't know why are there so many linear algebra books. They may have
different notations.

SW: What about eastern European books? DG has looked through eastern
European books.

BB: We should look at Japanese books.

PL: has some Japanese books. The notation is close to the American system.
Not the same, but close.

MUS: He has two Japanese books, and the math was in English.

DC: To make progress with writing the MathML 4, we should have a meeting
that does not discuss intent.

NS: that will be the topic for next week.