Bienkowski: Learning Registry Video Lecture | Understand Learning Registry: Online Teaching tool - IT & Software

welcome to the eighth and final week in
the learning analytics and knowledge
open online course this is actually
constantly also our final presentation
in this course as well we're
particularly pleased today to have
mehrabian Kowski join us I think I met
Maria the first time is about a year and
a half two years ago NSRI and where we
had an opportunity to have some
discussions around learning analytics
and educational data mining and she's
one of the authors of the influential
Department of Education report looking
specifically at educational data mining
and learning analytics and so she's very
informed on these topics and it's a
pleasure to have her involved here so
I'm not note Marie I'll pass it over to
you welcome
alright thanks George for that
introduction so what George asked me to
talk about today is a project that we've
been working on with funding from the US
Department of Education and the US
Department of Defense called the
learning registry and I did want to take
a minute to acknowledge not only the the
funders but also the people from the
open source community who were
contributing to the project a special
call out to the National Science digital
library folks the on the disk jay learn
experiment in the UK that did quite a
bit of work with the on piloting the
learning registry and setting up nodes
and then also a small company here in
california called navigation north that
did some you know served as early
testers for some of the technology so
now i'm going to try to go to the next
slide and I realize that must be that
okay so why are we interested in why
should we worry about learning resources
you hear a lot about open education
resources and I wanted to take a minute
to give you a little bit of context that
helps me get out of bed in the morning
when I'm thinking about thinking about
education so I think the tide is
actually turning and we're starting to
see that some of these traditional
education practices are starting to fade
this idea that you have to have these
hour-long classes where there's a seat
time requirement students are lectured
to they're grouped by age instead of
ability and that you've got this sort of
paper textbook as your primary Learning
Resource no cell phones and really
what's interests of interest to this
group is the kind of open feedback loop
that you find in some of these places so
what we think the trend is toward
hopefully is this idea of deeply digital
learning you know the flipped classrooms
is more engaging game based learning
badges that you get for competencies you
have learners collaborating by ability
and interest in what's relevant to them
and then this idea of the data it
ecosystems where you have these highly
instrumented environments including
leveraging the internet of learning
things so analytics and data mining I
think has a lot to offer in that space
as you've probably been talking about
for these past weeks really I'm seeing
it as a way to do continuous improvement
of courses curriculum and applications
through things like a be testing a lot
of people are starting to work on more
continuous testing and stealth testing
and some of the things that we're really
interested in for the learning registry
is this idea of recommending online
learning resources so you can build your
own pathways and I have a little
screenshot here from the newton's system
that shows a particular pathway that
this student took in
learning in this math area and then
another thing that we can focus on with
the these large amounts of data is
helping students regulate their own
learning getting the right challenge the
right encouragement and the right
engagement as they're going through
learning experiences so that's kind of
the big motivation for our work on the
learning registry and you'll see we'll
be getting down into the weeds a little
bit more but always trying to keep in
mind the end goal of the of the student
so what I wanted to also let you talk or
also talk a little bit about are some of
the problems that we've been seeing with
the ability to do personalized learning
and recommendation system so you have
these repositories a lot of people in
the digital library community have been
very active in building up repositories
but they can to be really stove-piped
and they're really not they're sharing
data through these pipes that they
connect with each to each other with
like oai PMH for example on these
harvesting protocols and what happens is
what what the observation has been is
that people don't update metadata they
don't care they don't they won't add it
at the outset they just dumped there
they link to their website and don't
really do any descriptive work about
what that website does for learning and
then we have this observation that
there's this data exhaust that's out
there people are doing things on the web
they're doing learning in classrooms and
all that exhaust is really not going
anywhere it's not powering any kind of a
future work in learning or improving
learning and so what you find is that
there are a lot of learning resources
out there and this is really why the
Department of EDD was interested in at
the outset and they're really these
learning resources are developed there
they're federally funded they're funded
by federal agencies in the United States
and they're just language languishing on
the cell the shelf because the data is
locked in and locked down and what we'd
like to think of instead is so instead
of having these stove pipe repositories
that share
hold data and share via these pipes that
we've got data that shared in the open
so people can aggregate it they can
amplify that data they can do analysis
of it all toward this goal of improving
the access to learning resources so
instead of having this kind of scale
metadata where people say we can't even
plan on goodnight as metadata at the
outset we want to think about enriching
this data pool this met its social this
metadata pool with social usage data
this is also called tension data
metadata or para data to really add
nuance to metadata and to start to
harvest some of the kinds of analytics
that we're seeing from sentiment
analysis for example and we instead of
wasting the data exhauster having it not
be powering anything we want to have it
collected to power the social life of a
learning resource so it can get a little
bit more a little bit better play for
learning and instead of having learning
resources languish on the shelf we want
to have them live in a rich ecosystem
where they are used people know how
they're used the developers know that
they are they can be improved and and of
course we can benefit personalized
learning in this case and then of course
we want to have democratization and
liberation of the data and setting
instead of it being locked in and locked
down so the learning registry was
conceived of as a way to put data into
the open for aggregation amplification
and analysis it's a distributed system
so it's a set of nodes that pass data
amongst them amongst each other and it
permits you to share metadata about
learning resources it was developed
through open-source process and we've
engaged a number of different community
members throughout the couple years that
we've been working on this we've been
talking with Learning Resource creators
so people who want to publicize their
open educational resources but also
publishers so people who want to be paid
for the work they've done on creating
educational
versus people who curate people who and
people who consume and want to give some
feedback into the system of improving
learning resource use so it's a story
forward data exchange network
technically that what it is what it is
and it's often difficult to explain to
folks what the learning registry is
because there isn't although we have a
web page there isn't really a place
where you go to to use the learning
registry or to log into the into the
learning registry it's not a search
engine it's not a repository it's a
registry of metadata information but
what we envision is that you can have
diverse user services be built on top of
the application interfaces that we've
created and then that use the data in
the learning registry so this is
addressing the problem of being able to
enrich the learning resource data with
para data or attention meta data to add
nuance to metadata so here's an example
of how the learning registry might work
so the middle column shows the learning
registry where you have metadata and
it's actually stored in a timeline so
you could do for example a trend
analysis from the data that's in there
and so we've got a couple of players
that we've used in this representative
scenario so the US Naval Air and Space
Administration NASA has some interest in
getting kids excited about stem science
and technology and so they've got the
Mars rover and they want to have some
learning resources about that so they
might go ahead and post a message to the
learning registry that says hey I've got
some metadata about this Mars rover
animation then the National Science
digital library which is a us-funded
digital library of STEM education
materials could say hey there's a
through the learning registry could
discover that so there's some discovery
that's happening here and say well that
looks really good I'm going to take that
resource find its net get its metadata
from the learning registry and then you
provide that resource to my to my users
but you could also have a teacher portal
and that's what this cge online
represents that's a website that
navigation North has created for career
and technical education so they know
that design thinking is very important
in career and technical education so
they want that Mars rover animation so
they pick that up from the learning
registry and then nsdl has a user facing
website so they can go ahead and have a
teacher that actually comes to their
website and that teacher can say oh I
like this uh Napa animation and that
information can be fed back into the
learning registry for later use a career
and technical education at edge teacher
can actually on say that that NASA
animation aligns to a standard a
curriculum or a learning standard and
then that information can be put back
into the into the learning registry so
you've got this timeline of metadata
descriptive metadata and social metadata
about learning resources that on then at
the bottom we show the cloud service
providers that can take that data they
can aggregate maybe life data that allow
teachers really to to share their
opinions because teachers are often
isolated especially in K through 12
which is kind of remind what Lamar I had
been a lot you might have alignment data
and I'll show a little bit of examples
of that later and use that to improve
searching so you can search on alignment
data for standards or maybe I'm doing
analytics on the social metadata as I
show on the right service provider that
just provides advanced recommendations
so this kind of a scenario a day in the
life maybe of a learning resource in as
its metadata
goes into the learning registry so the
next slide shows some of the way that we
can on reify this idea of data exhaust
we're trying to collect it to power the
social life of a learning resource so
this shows annotation allah kinematic
and pedagogical para data and this list
was created by the National Science
digital library interacting with
teachers and asking them what kinds of
things they thought would be useful to
know about learning resources so there's
a kind of a whole refinement of this
descriptive metadata where your rating
you're commenting you're voting there's
diffusion through user actions you're
favoriting you're downloading you're
sharing to social media so that's moving
it moving it around there's also
pedagogical para data where you're
actually looking at the educational
content and the utility of it may have
been lined to a standard is it embedded
in a learning resource so you can see
that if we knew this kind of thing about
learning resources and we could harvest
that how much more useful would it be
for a particular learning resource and
what we've observed is that you have
these popular learning resources like
the nasa mars rover happens to be very
popular and it shows up in a lot of
different places there isn't just one
repository where you find it and but but
there's no place to go and aggregate all
that information together so what we
want to do is use the learning registry
to just let resources live in this
Richard ecosystem where you can actually
look at something like the Mars rover
and have a summary this is a mock-up
this is not an actual an actual website
but you can imagine having the resource
information having usability ratings
accessibility ratings tags if people
have put on it voting schemes this kind
of shows the power of the on the data
that we could pull out from the learning
registry
mention one more example and then I
think I'll take a break in case there's
any questions so here's an example of
what this might look like and this
example especially of interest to
teachers because they often when they're
looking to put lesson plans together in
the K through 12 spaced are they're
trying to innovate in their practice
they're trying to find interesting
things they're trying to find things
that are online but this data is has has
not yet been liberated until we can pull
it out of the learning registry so this
is an example chrome plugin that we
built that actually amplifies the google
search so you see we did a google search
for interactive volume and surface area
and what happened is the plug-in queried
the learning registry to see whether
learning registry had any metadata or
para data about this particular resource
indexing on the URL and then it inserted
in Genesis this is legal I'm still with
the Google search page inserted into the
search results information that was
about this resource from the learning
registry and so this is kind of the
canonical example of how we see this the
data can be extracted from all these
different disparate sources put into
this data pool and then pull back out in
this useful way on where it's cutting
across really these different stove pipe
systems so I'm going to go ahead and
pause right here this is another example
of recommending resources based on
alignment data and I'm just going to
check the window here and I don't see
anyone in the chat so um I'll just type
in here in case somebody can see V any
questions thank you
you
okay a place we can get your slides okay
yeah okay so i can provide the slides i
have to realize I still have the floor
okay so on there's a very interesting
question about on the statement of the
problem and how we get the different
stove pipes to cooperate and there's
also a question about oer commons so it
is it is true that bootstrapping the
network effect is a tricky problem here
on the cooperation is really pretty
simple all they have to do is um really
a dump their data into the registry
there isn't any you say cooperation and
that maybe that's not the right way to
think about it because really what
they're doing is just sharing their data
now we have had some folks who feel like
maybe sharing their data is going to you
know give someone else a competi
competitive advantage over them but if
these are open educational resources and
we're all in the business of trying to
get them to you know be located so that
we can you know benefit education having
a cron job that runs once a night that
takes your updated data and puts it in
this learning registry isn't a big
burden on your system and it will help
on you know the curation that you've
done it will help those resources be
more findable it's true that it wouldn't
necessarily direct traffic back to your
site which is often what people want but
we're also hoping that on the idea that
you know people want to benefit
education and that should be a place
where we're sharing is what's going to
motivate people to participate so that
was the stone
those pipes will get you the slides so
oer commons is a little bit different
because that is a that is a repository
that is being curated by the by the iska
me folks who run oh we are commons and
they are actually publishing their
metadata into the learning registry but
what's interesting about oh we are
common is they do have a rating scheme
and a you know favorite and so on on but
you if you go to oer commons you'll see
the Mars rover if you go to nsdl you see
the Mars rover if you go to a National
Council of teachers of mathematics you
see the Mars rover and the point is that
getting them all to share data would
make an require an N way collaboration
and this way with putting it all in the
learning registry they can make it we
can make it all available so it's
similar in the sense of the kind of
social metadata we're collecting but not
in terms of the functionality the chrome
plugin um currently is I think it's in
the chrome store but I don't know if
we've updated it with some changes that
chrome has made so I can pass that out
to George so he can post it in the
course and then the standards open
systems right to to streamline access
and analysis of disparate stovepipe data
let's see if I have if I answer that as
we're going through the rest of the top
and then the if teachers do not care
about metadata it's actually not the
teachers that don't care about the
metadata interestingly enough it's the
providers of the learning resources and
the people who curate learning resources
that often have trouble giving metadata
in or have trouble on you know getting
the right metadata what's the right
grade level for this what's the amount
of time it should take for a student to
do it what kind of language skills does
it require so on
I think that what we're hoping is that
the pair data gives you a kind of more
up-to-date continuously up-to-date on
information so that they would start to
care about it because they see resources
that are more relevant kind of bubbling
to the top and I'll show some examples
of that when we talk about the alignment
data ok so it's I'm going to hope that
I'm answered questions and I'll try to
get to the the standards question as
we're moving along so I think we talked
this is what I've been talking about
where we've got these different learning
resource systems that are generating
this kind of data that can be connected
to the learning registry network and
we've got these publish and consume
api's you can work with learning object
repositories like oer commons and nsdl
teacher portals like mathematics portals
or um I think the National Archives has
a teacher portals for their your
learning resources search tools i showed
the chrome plugin learning management
systems we've been talking about hooking
up with them and also i mentioned here
instructional improvement systems
because some of the states here in the
US are interested in instructional
improvement and are building these large
data systems and are being asked to
participate in the learning registry as
part of their on improving their
practices not just within a state but
across state as well so that's pretty
exciting for us so I just wanted to give
you a little bit of a picture of on some
of the data this this is a couple years
old and I didn't have time unfortunately
to regenerate it so this gives you a
sense of what kind of data is in there
we have a lot of science data in there
because of the extraction that we've
been doing from the National Science
digital library and the some other stem
oriented sites but we've also got work
on literature a mayor
and democracy algebra algebra and
functions not algebra and fun even
though we know answer it and fun there
is some of the on UK data that is put in
here is kind of in the lower right-hand
corner so this gives you a picture of
what kind of data what kind of metadata
we have currently in the learning
registry so a little bit of technical
detail that's built on a document
distribution system called couchdb that
uses a no SQL database system for on
managing the the document data it is
metadata agnostic we're using a JSON key
value representation to encapsulate
metadata in any schema it doesn't mean
that we're translating meta data into
these JSON format it's just an
encapsulation so we put a little bit of
of writing around the metadata on that
the things that are specific to the
learning registry and then store the
metadata on in the in the learning
registry registry the social usage data
is modeled initially on this n LG n stl
comp era and also effort that was
started to allow people to kind of RSS
style broadcast their activity i'm using
activity streams on this I think was
before a Twitter was actually started on
the thing that's a little tricky then
I'll mention here on full disclosure is
that we're using the URL for the
resource we struggled a lot with how to
do on how to uniquely identify a
learning resource and the URL was what
we kept coming back to because one of
the design principles that the
Department of editors very insistent
upon was that there be no centralized
Authority or centralized registry it is
a distributed node system and they
want any single point of failure or
anything that's the department of EDD
would need to keep say funding after the
startup funding went away and then of
course we've got the problem of you know
people submitting things they shouldn't
be submitting with URLs to things that
people shouldn't be looking at if their
teachers and students so we have a
digital signature system that's actually
based on PGP pretty good privacy that
does submitter identity verification
with this digital on signature so I
wanted to talk a little bit about the
pair data schema because this is where
some of our leading-edge work is in on
this gives the schema with the Act this
is based on the activity streams actor
verb object and then what it was what
it's related to and what the the content
is some human readable string we're
using this for things like alignment
data number of occurrences start and
stop time on context of where the
learning resource was used with it used
in a second grade classroom was it used
in an informal learning environment and
we're finding that this is needs a
little bit of an extension so we're
actually working on on version 2 because
we need a little bit more description in
the actor field to describe more than
just a string description about the
actor because we want to make sure that
we we've talked about wanting to build
trust networks based on this and you
want to have a rich description of what
the actor is if you want to actually say
try to decide whether the in alignment
they did from to a standard a curriculum
standard has some weight if a teacher
does the alignment that gives you one
kind of feeling of how much you should
trust that alignment but if a district
curriculum person does it you have a lot
more
you have a lot more faith in that so
we're actually doing a refresh on this
based on some work we're doing on with
folks who really have a strong interest
in doing para data so that will be in
extending I want to talk a little bit
about briefly about data services this
is a kind of a broad brush overview of
the learning registry you know generally
we have these I show on the right the
kind of familiar to digital library
folks interfaces that we provide obtain
harvest the OA IP mhm we also have a
interface that we call slice that lets
you cut out whole chunks of the learning
registry based on things like on
alignment to a standard or a tag of a
particular type or on a submitter so you
could pull out all of the data that the
Smithsonian has submitted for example
but were and this shows the distribution
going through the learning registry
network and what we're trying to do is
build data services on top of that they
give you another level of abstraction
beyond the obtained and slice and
harvest on harvesting applications and
then the publish and sword on publishing
application so that's that's kind of the
obligatory slide for the digital library
community and it doesn't work okay and
so I wanted to just acknowledge and give
you some pointers to and it will be
helpful especially if when George posts
the slides that we do have some people
that are very active active adopters and
experiment or so navigation north and
you can go look at their common core lab
site in bloom in the what used to be
called the shared learning collaborative
is using it for their student on
integration systems to do to support the
infrastructure
or a search and a browse function for
digital resources there's a link to the
jailer an experiment that the that was
completed in the UK i think it was
University of Manchester and then the
state improvement systems is work in
progress so I don't have any links to
that yet so I wanted to take of a minute
we just did a proof-of-concept
internally in pulling some data out of
the learning registry that's letting on
that's doing recommendation by building
these on network graphs things weren't
necessarily social network graph that we
were doing these are just some examples
of what you could do with learning
registry data so let me jump in and show
so our example here again showing on the
kinds of data that we anticipate
especially for the state improvement
systems and for the in bloom work that's
going to be really important in in the
United States we had for a long time on
separate states each had their own
standards so Utah had standards
California Florida and they had spent a
lot of time on aligning their resources
to standards different learning
resources to these standards and now we
have this new thing that's come in where
a state consortium has agreed to adopt a
set of common standards for mathematical
mathematics and English language arts
these are the common core standards and
so what we what we believe the learning
registry can do and why the state
improvement systems are using it is that
they can help locate resources by doing
cross walks among the on state standards
there's you tomm a standard x and if
it's the same as this Common Core
standard then you give a teacher in Utah
knows that they can use that learning
resource accounting 250 to match that or
to help a student
that math standard that math standard x
so for a while as teachers are getting
used to these common core standards they
can still kind of fall back on their old
familiar state standards this also shows
you know some favoriting and rating data
as well and then here's a more complex
on more complex aggregation that we've
done with we've actually built this in a
graph database called neo4j Jim Crow was
the lead architect and lead designer of
this and you can see that we have a much
more complicated situation but it also
provides us with more opportunity to do
things like collaborative filtering so
if i have if I'm an educator and I
favorite a learning resource and I know
that learning resource is a line to a
standard I might want to see other
learning resources that are also aligned
to that standard and if two educators
both take some action either rating or
favoriting a particular learning
resource we might want to do a
collaborative filtering application
where we share with educator a what
educator be has been looking at and
shared with being what a has been
looking at and this gets us to the
canonical on you know amazon.com uh
space and that everybody uses a
canonical example of collaborative
filtering so we've actually built this
in a neo4j with an extract of learning
registry data and we don't have a you
know big user interface on it so it's
not something I can I can demo but I do
have a pointer to a forthcoming book
chapter we have about this so what's
next so I'll wind up here and leave some
time for questions we are working on
sorry that got cut off we are working on
more recommendations by clustering users
based on their actions on we still have
work to do on identity on using URL is
not quite working we're having to do
assertions that a URL let's say oer
commons is pointing to the same URL that
some place like the Smithsonian we're
still trying to bootstrap and get this
network effect to really take off I
mentioned the fear of lots of
competitive advantage we haven't done
performance statistics on the neo4j work
too so that's something we're we're
starting up now we are we did say we're
metadata agnostic but we are trying to
tip our preference to the learning
resource metadata initiative which we
see is having a lot of promise because
of its adoption by schema.org and the
potential for micro tagging to let
search engines pick up learning
resources but we think even if search
engines are able to find learning
resources as easily as they find recipes
we're still going to need the learning
registry because they're not going to be
aggregating on social metadata for
resources there's also a pointer i gave
to an interesting extension that we're
doing with the company a nonprofit
called Benetech which is looking at
accessibility metadata and if you go to
that wiki page you'll see the proposal
that they're making to schema.org that's
aligned with the learning resource
metadata initiative for metadata to mark
up resources that are accessible to
students that have those needs and we're
also on starting an effort when we're
doing our pair data refresh to put
accessibility para data to find out what
people want to say about resources that
are accessibility in the social metadata
sense so that's it on these are
references of the
the book chapter the book should be out
in a couple of months on we have a
educational technology magazine if
you're interested in more information
about the work that just did on and you
want to see some scenarios of how they
use the learning registry that towards
network knowledge paper is a good one we
have our lak paper from last year and
then lastly I list the github site for
the on for the learning registry if in
case you want to go check out the code
and documentation and then finally I
just wanted to make sure I'd be able to
call out to the folks who provided these
slides and also the people who are the
main github committers Jim khlo from fri
my company and then on Walt grata who's
with the advanced distributed learning
lab at the Department of Defense so I
have a few minutes for questions
you
ok so the application that allows you to
browse learning registry data on we've
got a couple that need to be refreshed
with the new the new you can't really
browse it let me say that what we can do
is we can pull out aggregate statistics
like the box diagram I showed what I
think the in bloom initiative is
probably I'm going to go out on a limb
here and say within a couple of months
of having a browsing application browse
and search application on top of
learning registry data and then an ideal
outcome for the impact of the learning
registry once it's done is really in you
know a my best scenario is one in which
you know people agree that for at least
open educational resources it's
important not to just try to drive
traffic to your site but to make the
resources discoverable and that we can
support teachers in putting innovative
learning resources into their classrooms
and that students who want to do on kind
of self you know self-paced learning or
explore more things on their own can
find these resources that are
appropriate for them so it's really all
about it's really all about the sharing
and that would be the ideal outcome is
that people are continuously sharing
this information you know I'm a big fan
of the bitly a website and I always hear
talks from Hilary Mason about how she
doing interesting analytics because of
what people bookmark on bit ly but all
that information is inside bit ly no one
else to my knowledge you know has access
to that so it would be wonderful if they
would open up anything that they thought
thought was educationally oriented and
put that information into the learning
registry for everyone to use
so the lr on so the question is the lr
lrm I schema supported by the learning
registry we support all schemas what
we're currently doing now when I said
we're going to tip the the tip the
balance toward lrm I we're creating a
schema validation tool that will let you
publish lrm I schema formatted metadata
into the learning registry and we will
validate it because the the in bloom
search and browse functions want to have
or they depend on having a well-formed
lrm I data and so we're going to build
that on that validation format for them
the validation tool for them okay so the
quality of discoverable resources is an
interesting question and it's that maybe
I should have said that that was on my
ideal outcome for the impact of the
learning registry now but thanks for
that plant John really what we'd like to
have is the have just to discover what
makes on quality how we could define
quality you know we think it would be by
usage but that's hard to do unless
you're putting it into a learning
management system you know if a teacher
just sends their kids to it you know you
get all of a sudden you get you know 50
hour 30 hips you know that might mean
that you've got 30 kids in a classroom
that are actually looking at that um we
also think that ratings are important on
peer review is a little bit more of an
involved process and I know oh we are
Commons has a quality rubric that
they've been using that will let that
will let teachers do a more structured
review of it I think the point is that
it's to be discovered and right now it's
there's no way to get that information
out and we're hoping at least to provide
on some kind of forum for people to give
that sort of feedback not only for
quality but we're also we also think a
the accessibility
para data will be another another thing
that's of interest to a lot of people
there it's pretty high stakes if you go
to a resource and it's not doesn't have
closed captioning it doesn't have a font
that you can use on you know things like
that so the common core alignment okay
so let me see does the common core
alignment among states I should be
repeating the question provide how the
standards for each course in each
state's are equivalent so the common
core alignment is really on aligning a
standard to a resource and what you do
to say that a standards are equivalent
is something called a crosswalk and the
people who are doing those crosswalks
are professional curriculum developers
and then we're taking those crosswalks
and putting them into the learning
registry for people to use so there's
choose me there's another question about
Drupal and I think the differentiation
between the learning registry and Drupal
is that the Drupal is a content
management system and this is up mate
lets me make an important point learning
registry is not a content management
system it's a metadata management system
which is why we've probably called it a
registry because it's not the learning
repository so we are not dealing with
any Content the learning content is
hosted on the on the sights on that
creative or where the URL points so
we're not doing any kind of harvesting
of of content okay so on slice as a
stream feed API I would recommend that
for that question because we are running
out of time on what slice does is it
does do a stream feed but it doesn't do
it on a continual basis so you can't set
up a recurring feed that does that
filter
like an RSS feed it just does a one-time
delivery of the sliced resources so we
have other API filtering api's that
would let you do that kind of more RSS
like on a stream feed okay learning
registry in tin can is another good
question on tin can the folks who worked
on the para data specification and
activity streams on we're also the folks
who were working on the tin can api and
you can submit tin cans statements to
the learning registry um i don't i
haven't looked at their schema in a
while so i don't know how harmonious it
is but it's a very good reminder that i
do need to look at that especially as we
do our para day to refresh activity
okay i'm going to send it back to to
George because I think just running out
of time and I'll need a little bit of
transition time here too well thank you
Mary I think I was very informative talk
and certainly given the increased
attention now to a variety of ways to
try and provide academics with a better
level of analytics whether it's through
a distributed system such as learning
registry or whether it is through some
of the activity that's now going on out
of the drawing a blank here the activity
that's happening with the LTI initiative
with IMS tin cans so it's interesting to
see just the amount of activity now in
trying to gain better opportunities for
data to be shared between systems across
silos but also to add para data layers
that gives greater insight once we get
down to the X actual analytics activity
so thanks very much for you safe travels
and take care
you