Tip:
Highlight text to annotate it
X
>> Melissa Terras: Thank you very much
and it's a delight to be here.
Thank you for the invitation to come and speak about our work.
Most of my research over the post 10 or 15 years has been
in imaging; how to image cultural and heritage material,
how to deliver it to people, and then wondering who's looking
at it once we put it out there.
For the past 40 years, most libraries, archives
and museums have been engaged
in how can we use computational technology
to deliver some things to our user community.
So this predates the Internet.
Back in the 1960's - 1970's, people were already thinking
about how to digitize stuff and deliver it to people.
Of course, given the world-wide web, most museums, libraries,
archives have a website and there is an expectation nowadays
that you would put your collection online.
With that comes the rhetoric
that it increases access to the collection.
Searches in science will begin, people will be able to do things
that we haven't done before
if we can access the collection online.
You wouldn't have to travel to New York
to see this one document which is over there
in the New York Public Library.
You can sit at home on your laptop
or your sofa and access it.
You can also reunify collections which have been dispersed
and put them all together.
This is all fantastic and we love digitization.
We work on how to do it properly and how
to use new imaging technology, specifically,
how to read damaged and deteriorated things.
But also, we're very interested in the whole aspect
of digitization; how you do it properly and who is using it.
This question, though, of who is using it has become very --
it has become very important these days
because of the economic climate.
Just like anything else, funding bodies, especially the U.K.
and across Europe, don't have as much money as they did
in the 1990's to put into digitization initiatives.
And now governments and funding bodies are asking well,
where's the proof?
Where's the proof that this has actually changed scholarship?
Where's the proof that people have actually been using
this stuff?
Most libraries, archives
and museums have not been forthcoming or very open
about who is using their online collections
and why they're using them, and most of them don't know.
They don't have the money or the time to investigate.
So now they've been asked to suddenly prove this investment
for 40 years' worth of digitization
and we don't really know who's been using it
or what we can prove so that we can fund more digitization
for the future.
So what we're interested in is who is coming
to use these collections and what are they doing with them
when they actually get there?
The British Museum collection database online is a really
great case study for this.
As Matthew said, it was launched in 2007.
By the end of the 2009, there were two million objects online,
over 600,000 of those had one
or more images associated with them.
At the moment there are 670,000 objects
which have one or more images.
So it's an increasing amount of things which are going online --
increasing amount of images which are being put online
and this demonstrates a phenomenal investment in time
and energy and effort and infrastructure to do this
over a number of years.
So I do not come here to bury the British Museum,
I come here to praise it for the effort and then to look
at who is using this collection?
What do we know or what can we know and how can we find
out who is using this collection online?
We have some experience at UCL University College London Centre
for Digital Humanities doing user studies, specifically,
in a cultural and heritage sector.
We've done some local analysis of Internet resources
and the arts and humanities and we'll talk about those analysis
in a minute and what that actually means.
We've worked with digital libraries with people --
experts in human-computer interaction
to see how people are using digitalizing
and how we can make them better for people to use.
We've done some research with archeologists in the field
to look at how we can take technology into the trench
and speed up the archeological process.
We've got an ongoing project now called QRator
at the Grant Museum at UCL which won a big award a couple
of weeks ago where we are putting QRcodes and iPods
into the museum and seeing how we can get people
to engage more with museum content.
I have a raft of Ph.D. students working
with the British Library, the British Museum,
the Science Museum, the Peach Museum, the Grant Museum,
all looking at user studies of digital material
and digital outreach and how people are using them
and what they're doing it with.
This particular study I want to talk about was done
in combination with a research assistant
of mine called Claire Ross, and a work placement student,
one of our master students, who came here for six weeks
over the summer to do some work with Matthew and his colleagues.
And we decided to do a study looking
at who was actually using the database collection online.
Now how this kind of came about is we had Claire to work
on another project but she had a few weeks
where she wasn't doing much so we used her time wisely
and were able to do a little bit extra
than we said we're actually going to work.
So we were based on project called Linksphere where she got
to come down to the British Museum
and help out with that study.
So we want to study how people are using collections online.
How would you do that?
There are two ways to do it; one, quantitative
so you can count people, and one,
qualitative so you can ask the kind of vague reasons
about why they would do things.
Why they would come and do things.
What they are using the things for it, stuff that you kind
of need a sentence to answer for.
And to do quantitative analyses of this kind of thing,
you need to look at server logs.
What are server logs?
Every time you access a website, you leave a digital trace
of the fact that you sent a request to that server
to send you back some stuff.
And if you set up your server properly,
the server keeps a list of every single request it's had
for a text file or for an image or for a movie
or for a [inaudible] model,
anything that's sitting on that server.
So the server generates these huge files of people from all
over the world who are coming to look at this.
We can also tell how long people have been there from.
We can tell where they came from.
It depends on the setup
of people's individual computers we might be able to tell
if something came from Cambridge.
We might be able to tell if someone came
from a specific lab at UCL.
Sometimes we just know the country.
So it's not an exact science.
In general, we don't know the name
of the person sitting behind the computer.
We also don't know if that person got up
and left the computer and someone else sat there straight
after them and looked at the same website.
So it's not exact; we have to interpolate between them.
But we can look at logs, we can look at links as in
which websites point to each other
or we can use commonly available tools now
such as Google Analytics to actually see and analyze.
It's very easy from our own desktops.
The qualitative means that this is a kind thing we do
information studies at UCL where we do a lot of surveys,
interviews, focus groups to try and uncover the stuff
which the quantitative stuff just can't tell you.
You might know that you're getting a lot of people
from New York coming to look at your website,
but you will never know why.
And asking them this question is the thing that that you need
to do to uncover that.
If you're interested in this at all,
there is a fantastic resource called The Toolkit
for the Impact of Digitised Scholarly Resources,
which was developed -- funded by the JISC Funding Council
and is available from the Oxford Internet Institute
and they run you through how to do all of this type of stuff
if you want to evaluate an online collection
or online digital resource.
So I would say that's the first place
to start if you're interested.
So here we have the kind of things that we can do.
If we use Google Analytics as the way we're going
to look who's coming through the website,
we can find out where people are from.
We can know how long they are on the website,
how many pages they were visiting at the time.
We can tell the traffic --
that is do they come in through a Google search
or from a Yahoo search or do they come from a different link?
Who's pointing traffic to your website?
We can do online surveys to get some in-depth answers
from people but they have to elect to apply to the survey.
And we can also undertake in-depth interviews
so that we get one-to-one responses from people.
And this study, because it was only six weeks' time,
we did not have the time to do in-depth interviews
with a whole range of users and we'd like to do
that in the future, contacting a whole bank
of people we haven't contacted just now and get back in touch
with them and asking them more
about their own personal use of the collections.
So context and the last calendar year,
there were 10.5 million visits to the British Museum online.
There were, however, 60 million page views,
which means on average that every person
that comes is looking at five web pages.
30% of these people were looking at the visiting
and what's on sections.
So 30% of people were saying well, "What's on?"
and "What time is the British Museum open?"
And that's natural, right?
If you think about how you use a museum webpage mostly a lot
of the time looking for access information; how to get there,
when it's open that kind of thing.
Of the 10 million visits, 1.2 million included a visit
to the research section.
So there's a lot of people coming to the research section
which contains the collection online.
But when they got there they were responsible
for 29% of all pages.
So a small number of people are coming to the research section
but they're looking at many,
many more pages once they get there.
So they're coming there to do something.
Just to pause for a minute and think how this relates
to the physical visits to the museum.
In the same time period there were 5.8 million physical visits
to the museum so there double than might of virtual visits
than are physical visitors.
And what is the relationship of the physical museum
to the virtual museum?
That's something we're still grappling with.
What should the virtual museum be in relation to the --
and is there a connection?
And should we make that connection clear?
It's something that people in museum studies have been talking
about for a long time.
The study that we did was looking at stats from 18th
of June, 2009 to 17th of June, 2010.
We looked at over 8 million visits in total.
On average, people looked at six pages.
Thirty percent of people returned to the site
so there's 70% new users every single time.
230 countries came and nearly a million people searched
for something on the webpage but when they were searching
on the general search they were saying things
like Rosetta Stone Egypt Mummy.
This is not the stuff that we're really interested in.
We're interested in when people go to another level and look
at the collection database and are doing more specific searches
for things rather than where is the Rosetta Stone in the museum.
In that time, there were 37 individual searches
of the collection's database.
So that is what we are looking for.
And most people, when they search the database they then
spend a minute on the site.
So they're doing something; they search, they get some results
and they're reading them.
And 30% of these people are returning.
Do remember that figure; we have 30% returning people
who come back and use the database over and over again.
When you look at the stats, you get a lot
of really dry data that comes out.
You get this kind of thing;
a chart which is, you know, fantastic.
You can see that when people are coming to search the database,
60% of people are coming through Google.
So they're asking how to search the database through --
type it into Google and get directed to the museum.
The interesting thing for us is 21%
of people are coming directly to search the database online.
So they've got a bookmark somewhere on their computer
that they know where it is and that should be interesting
because that means that people are taking it seriously
and have it and they use it regularly at home.
And another interesting thing about this is 15% of people
who are coming to look at the collection database online,
come from somewhere other.
So we looked at where people were coming from.
They come from websites like, not surprisingly, Wikipedia,
but also there's a lot of people coming
from the British Museum Shop Online, so they see something
in the shop and they are coming to search
for it in an actual catalog.
So that's kind of interesting how there's a link there.
So there's a very prominent link in the British Museum Shop
and they're coming back into the collection to search for it.
You can see some interesting social things as well.
This is how many more people looked
at the British Museum website
on a mobile device during that period.
There are absolutely no hits until the middle of October --
towards the end of October.
Why would that be?
Because iPhone came out in the middle of October.
So we can see suddenly but no one can get online
with your mobile device and they said well wait there you go
and that's going up and up and up.
People are starting to really access the museum
through a mobile device.
You can also see things like -- does this have a pointer on it?
Sorry. This thing?
Oh here it is.
Here, possibly a server outage.
No one can see the museum and then here.
Anyone got any idea what that is?
It's after November the 25th.
That's Christmas Day.
People don't search the collections database
on Christmas Day.
Who would have thought?
So you can see some various trends and things as true.
And one of our jobs is to go through and sift all that.
If you're interested in the analysis of this,
we do have a paper that was published on museums on the web.
You can go and see some of the more charts;
some of the more statistical stuff that came out.
We then put a survey online for a month
which targeted every second and fifth user that was coming
to the webpage to ask them about their motivation for coming
and looking at the collections online.
We got two and a half thousand responses.
That's amazing.
I mean, no one answers web surveys
so that's a fantastic response rate and really allows us
to talk confidently about what people are saying.
Only half of those people completed it
because it was 30 questions long,
so we did have quite a drop off rate towards the end,
but that was okay.
We knew that was going to happen.
So we had frontloaded the kind of questions we needed and know
that we were getting more and more kind
of amorphous questions towards the end, which were interesting
for us but people are not necessarily going
to actually carry out.
So we had 30 questions --
various ways to do it and we also had four defined tasks
that we sent people to find stuff
in the collections database
to see how they were actually using the search facility.
So what I'm going to do now is present some of the results.
Who is the using the British Museum database online
and why are they using it?
These charts are a bit dry so to help you
out I've put a gold star towards the top answer
and the silver stars toward the second answer.
So most people using the British Museum collection online are 21
to 30 years old.
So it's not school kids, right?
These are slightly older groups that are using
by at the least the people who answered the survey and above.
So we've got a whole spread of people there
who are using the collection database online.
29% of people who are using it are from the U.K. 91%
of those were from England, 6% from Scotland, 3% from Wales.
Most of the people from England were actually in London
so most visitors
of the collection database are actually based in London
when they do their search.
17.6% of people are from the States
and then we have the traditional long tail with Germany, Italy,
France, Australia and then a long tail
of countries following after.
We asked people how they heard about the collection online.
And this was interesting for us because most people had heard
about it from a colleague.
It wasn't from a Google search,
it wasn't from someone had told them about it;
that they should be using it.
So word of mouth, especially
in the scholarly community is really important still.
That's kind of interesting to know that it's not just
about upping your digital presence, it's also making sure
that you hit the right networks
to get your collection used by people.
Academic environments are also very important,
so academic staff filled with students, etcetera,
libraries within institutions, or --
and then from a link on the museum's website itself.
So the academic network is as important at pointing
to the collection as the museum's website.
We asked people what was the reason
that they were using the collection database online?
50% of people were using it for academic research.
Now this kind of proves that thing of like, yes,
if you put a collection online, people will use it
to do bona fide research.
We never had statistics before to actually show that
and we can show that of these people that answered the survey,
50% of the people were coming here specifically
to do some research.
Other people were doing it
on non-academic professional research, on personal interest
but that's a really a whole spread of reasons why,
but at the same time, it's good to know that people are using it
for core research for getting bona fide information
about objects.
We asked them, if they were in academic,
what kind of academic were they?
Most people using the database were post graduate students.
And you might go, oh,
post graduate students are just students,
but actually post graduate students are doing really
cutting edge research.
These are the people who are really digging around and trying
to find new stuff, whether they're masters
or whether they're Ph.Ds.
So this shows that the database has been used
for really novel research.
Following that is professors that use it.
So it's quite serious research the people
who are using the database for.
We asked people how do you expect to be able to search.
And unsurprisingly, a lot of people wanted to search
by free text, exactly the same type of search
as you would have in Google.
That's because people learn about how to behave on websites
from other websites before they come to yours.
So they've learned how to search on Google and they expect
to be able to search in exactly the same way when they come
to the British Museum.
But people also want to look at the type of object.
So it might be a period or vases something like that.
They want to be able to search in that way so they're looking
for what we would call
in information studies a faceted search
where they can choose different ways to go
through the collection and to sift it out.
They're also interested in the date; not so interested
in things like museum gallery.
That's not of interest to most people, which again,
shows this interesting link between the physical gallery
and the website and whether or not we should be thinking
about laying out websites in the way
that reflect the survey of the collection.
When we asked people what they were looking
for they said a specific object.
That means that most people know what they are looking
for before they come and they want
to find more information about it.
So they're using it to verify information
and to check information.
We asked people the types
of objects they search for on this visit.
And this shows an interesting quirk in the language.
What type of object?
And we got a whole range of answers from vases to Italian
and so it shows the different ways
that people perceive what you mean by type.
But there's an interesting link there
that most people are looking for prints.
Most people want to see prints
that they can't easily get access to.
We asked people how often do you use the collection database
online and 30% of them use it every day.
So they've got some really keen users there.
Only 27% of people said that this was the first time
that they are using the collection online.
Now, that should ring a little bit of alarm bells
because previously, I said that we have 30% of people
who are returners but here we're saying only about 30%
of people are using it for the first time.
So it should be the other way around, right?
So that means that this is a self-selecting survey --
that it's people that come back all the time just
to answer the questions.
So, therefore, the results are a wee bit skewed
and we have to acknowledge that.
But that's okay.
This kind of research you're never going
to get perfect answers.
You just have to acknowledge the statistics that you've got.
And we acknowledge that it's keen people
that use the collections all the time that answered that
but then it can be interesting for us
to get some results from that.
We asked them some opinions about the collection online.
They said it was easy to find.
They said that the layout on the collection was appropriate.
They said that it was easy to navigate
but some people slightly agreed with the fact
that it was easy to navigate.
So we asked them why and they gave us some improvements
which I'll go through later.
And most people wanted to come back and visit again.
So this is all fantastic stuff and it shows
that it's a well-used, well-loved resource for people
which is great to be able to articulate that for the people
that have put all their investment into online.
We asked again about this thing about physical location
of objects and people wanted
to learn a little bit more about that.
When we asked about improvements,
the big thing was people wanted to see more images.
Now this isn't terribly surprising; people love images.
But it takes a long time to scan in images,
so that's another issue.
It will just take time to scan in images or take photographs
of images or of all these objects.
People would like to see more images, more detailed records
and more things about provenance and 360 degree views
of objects, that kind of thing.
So the improvements people want were more images,
high resolution images, zooming enlarging images,
less steps for image retrieval once you've logged on,
option to return to the search piece
by the way people were going
through the website they wanted to change.
They also wanted more objects.
Not all of the objects in the museum are actually online.
So keep adding more objects, provide greater description
of objects, and provide references.
There's a whole list of things we could then give
to web teamers.
This is what people would want
if you wanted to make any changes.
We asked about social media and if people wanted to link
up the database to social media.
Over 65% of people said they did not wish
to use social media linked to the database.
They were not interested in linking
up to Facebook or Twitter.
This is somebody that's coming to do serious research
and not terribly interested in linking it
up so that's something you can say,
"Well we'll put social media to one side for a second."
We asked if people were going to reuse images [inaudible] and 90%
of people said, "I wasn't looking for images."
Fair enough.
Nearly half of them are not going to reuse it
but there's a lot of these going and we don't know what people do
with images once they got them from museum collections.
We don't know what they're doing with those images next.
That's a huge research question.
What are they doing next with it?
We asked them if they were on the British Museum's website
as a result of a visit to the museum.
And there isn't really a correlation between that.
It's a separate thing and people
in the comments said very poignant things
about the difference between visiting the museum
and actually looking
at the collection's database to do research.
And then we sent them some tasks.
I put up a picture of a pot.
I know this pot very well from my undergraduate days
and I asked the following question,
so just think about this.
You're searching for a great vase which you know is
in the British Museum as you have seen it in a print catalog.
It's an Attic White Figured Lekythos from around 490 BC,
which depicts the myth of psychostasia,
the weighing of souls.
The print catalog gives the reference B639.
What would you type in the search box to find the object
that you're looking for?
Come on. Give me some --
what would you type in if you're looking for that?
>> B639
>> Melissa Terras: B639.
Okay. If you typed in "B639" you would get 48 people who did that
and you would get 14 search results,
all of which are incorrect, and you wouldn't find the pot
that you were looking for.
If you typed in "psychostasia", 23 people tried that,
you would get no results and you wouldn't find the pot you were
looking for.
And if you typed in "Lekythos", 19 tried that,
again, you wouldn't find it.
So that's a problem, right?
174 people tried this task and only six people could find
that pot from that very detailed description.
So this is the kind of thing we heard back and went, you know,
this is a bit of an issue.
The problem is it's not B space 639, it's B639 all compact
that you have to put together.
So it's something we can pretty much straight away go, you know,
there's some tweaks you might want to make
to your search algorithm.
The same thing for this scroll.
You're searching for a hanging scroll with mountain landscape
which is seen in the British Museum's print catalogs
and ink painting and paper; 16th century,
attributed to Kei Shkei.
The print catalog gives the reference
"Japanese Painting ADD387."
What are you going to search for if you're searching for this?
What words you going to use?
>> Kei Shkei.
>> Melissa Terras: Yes, most people did that and that brings
up the correct results.
So there are things about the database
that kind of need to be tested.
I mean expect that when you do this kind of testing to be able
to feedback concrete examples of the kind
of tweaks people should be making to their collection.
You can also compare at the end of all of this some of the stats
that you get from the log analysis
versus the stats from the survey.
So on the left hand side here, we have the survey responses
and the number of the people that actually did the survey.
So 395 people were from the U.K. -- so most amount of people were
from the U.K. and on the right hand side we have the stats
from Google of the number of visitors that come
to the British Museum and come to their collection online page.
So, statistically, we have the same kind of countries;
United Kingdom, United States, France all over again.
So this is nice for us because it suggests to us
that we are getting a good spread of users from all
around the world in about the same way
as we are getting from logs.
So we can say, therefore,
okay we've got keen users answering the survey
but it's vaguely representative
of the community who are using it.
So these results stand.
So what are our conclusions from all of this?
I believe that this type of work enhances our understanding
and awareness of how scholars are using digitised collections
and, in particular, the British Museum's
information environment.
We have shown from the simple survey and log analysis
that digital resources are used extensively by academics as part
of their research practices.
We've got proof of all the things people have been saying
all along.
We have shown that collections
with a strong visual element are particularly useful.
People love images and they love to look at actual things.
We have shown a distinction
between a physical and online visit.
The physical visit is a leisure activity,
while the online visit is for research and information value,
so it's completely different things.
We have shown that social media isn't always the answer
to everything and that people don't always want all the bells
and whistles in being able to share everything
on Facebook and Twitter.
They actually want to come down
and do some work for some things.
We have shown that academics displays specific
information-seeking behavior, so we did a special analysis
on them and showed that they look for a known object,
they use specific search terms,
they show a specific goal or intent.
They're really keen academics.
They're really keen to find what they actually want
to find in the database.
And this kind of study gives us a bit of understanding
of search patterns for people and information-seeking behavior
so we can then develop databases in a way
that people expect to use them.
Make their databases better, make their searches better,
and make the content that we deliver to people better.
It's a valuable guide for further development
and refinement of museum online collections.
It shows a couple of areas of improvement
but mostly demonstrates that why would anyone bother to come
and visit the British Museum collection online.
It is very well-used and it is very well-liked
which is a really great thing to put back to the web team.
What's next?
We would like to do some longer research and how things change
over time because this research was done two years ago.
We did plan to follow up last summer, but I went
and had twins instead, so we couldn't do that.
But we do plan in the future to go back and look
at how this is changing over time
and I hope people's needs are changing over time especially
as IT skills are changing over time as well and access
to the Internet is changing.
We just put a major bid in to one of the funding councils
to do a two-year long study with another place at the Port
of Call, the National Gallery and the British Museum
and the National Library of Wales and Peach Museum at UCL
so we're waiting to hear if that's successful
so we can track and trace users in a much more detailed way
over a two-year period to see and give concrete evidence
of the value and impact of digitised resources.
And all I have to say to conclude to all
that is thank you very much to my student, David,
at the British Museum.
Thanks also to Claire Ross, my research assistant,
who's now my Ph.D. student at UCL,
and Vera who has now gone off to do fantastic things
at the BBC in the digital realm.
I also have to thank Matt Novak who's known
as paleofuture on Twitter.
He allowed us to use this image.
He put this -- I saw this fly by on Twitter.
He scans in old comics that have visions of the future
and this is a real comic from 1962 which says,
"Researchers thousands of miles away may consult books
in the Library of Congress or the British Museum."
We know that they may do it, but we've shown
that they're actively doing it for bona fide reasons
and they really enjoy it when they do.
Thank you very much.
[ Applause ]
>> Matthew: Just before we ask questions I was just going
to say that we are in the process
of redesigning the interface to our collection online database
which should be online later this summer.
And a lot of the stuff we got from the surveys
and the UCL's work is going to fit into that and they're going
to be some of the first people who see the [inaudible]
of action and help kind of rip it apart and tell us
where we've done it wrong.
And so thank you very much, Melissa.
I'm at this point quite interested
in hearing from our audiences.
Is there any questions from the audience?
Up here.
>> First question; how do we get some of your research teams
to come to our museum and do this for us?
>> Melissa Terras: Where are you?
>> Wellcome Collection.
>> Melissa Terras:[Inaudible] and you should talk
to [inaudible] there because she organizes student
work placement.
>> Oh, okay.
>> Melissa Terras: We have about 25 master students at UCLDH;
another 25 electronic communication
and publication students.
In fact, I know, because we've just run some stats,
that we have 185 students this year from our department,
all at master's level and on work placement
to 164 different institutions around Bloomsbury.
So yes, we're always looking for more.
We're always looking for more, so, yes, talk to [Inaudible].
>> Thank you.
Serious question.
How -- I'm interested to the extent in which this kind
of concentrated on scholarly use as demonstrating the kind
of usefulness of the collections -- the scholarly uses.
How would you go about kind of looking at its usefulness
to non-scholarly uses?
Would you do -- kind of concentrate on the segments
that are in the same picture or would you start
out the research in a different way?
>> Melissa Terras: I think it's a different suite of questions
that you want to be asking.
We very much did, "Are you an academic?
If so, go on and answer all of these questions."
And if not, let's do something else.
But the first pass, we wanted to focus on a certain audience
and we did certainly plan to go back, and we do plan to go back,
and look at a much wider audience because evidence
of impact and outreach is becoming much more important,
especially with digitization.
But also, I know I just presented about scholarly use
but it's wrong to think that there's not bona fide research
that doesn't happen outside the university.
And we are very involved
with a project called Transcribe Bentham,
and about crowdsourcing, and how we can crowdsource transcripts
of historical documents
and we have got fantastic volunteer labor doing fantastic
work, so none of these people are people at university.
So there's a kind of false distinction;
just because you work at university doesn't mean
that you're a good research and because you don't work
at university doesn't mean that you're incapable of research.
So there's an issue there that we have to deal with.
But you're right.
We're hampered by the fact that the only way to get in touch
with people is -- or know what people are doing is
through the logs or through the surveys.
And it's all we have to track off that.
And then collecting people's information
and then you do the focus groups and then you do the interviews.
So we're always hampered by that and there's always going
to be a slight disconnect about how many people we can get
in touch with and how well we can outreach to people.
But I think the response rate we got for this survey
over a month was phenomenal.
So people are interested in contributing back
to the resource that they're using and we try to reach
out to people in the same way but just
with a different suite of questions.
>> Matthew: We've actually [inaudible] askew on general --
on [inaudible] surveys but a majority of people who --
or lots of people who apply are collection online users
because you can see in the free comment field it's, you know,
"Have you got anything else to say?"
and it [inaudible] but it was merely
about a collection online and -- Tanya.
>> Tanya: I have a very deep voice
at the moment following a thyroid operation
but I wanted to say a few things.
First, it's true that the design
of the collection online interface is going
to be in print.
One of the reasons I think people aren't finding what they
want and the examples you gave, you know, the Lekythos
and [inaudible] examples are good ones is
because they're not using advance search.
The advance search isn't terribly friendly at the moment.
You often have to explain to people how to use it.
However, what the adults do, and what you haven't acknowledged
at all, unfortunately, is that one of the reasons, I think,
researchers find the database useful is
because we have developed incredibly good terminologies,
that's writing, names, and a lot of things are controlled.
In the case of a name, for example, it's a simple example.
There can be various ways of spelling a name.
And since we have names from all over the world,
this is terribly important that you do use the advance search
where it will pick up an alternative names.
Now because the advance search is being improved,
I'm hoping that you'll see a very different result
as to why people aren't using it now
and perhaps will use it later.
But the level of results is really determined
by which kind of search you use.
And for place names, for types of materials,
for types of objects, people often go
for higher level things like metal or china.
The word "china" is a country.
If you do a free text search for china, you're going to get china
in all sorts of context.
I don't think you stressed this side at all.
The other thing I'd like question slightly is, you know,
your analysis at the conclusion
that the physical visit is leisure
but this sort of visit is research.
You haven't mentioned the fact that there are loads
of student rooms in the British Museum.
A lot of the visits are academic visits and they do come and look
at the objects in those student rooms.
And you have to balance, I think,
the type of visits you're referring to.
And another thing I want to say is that, you know,
we have got an enormous collection
where people say they want more objects.
I think it's quite useful to understand
that you got too many records.
The collections varies [inaudible] six
or seven million.
A lot of the remaining objects are archeological chards
or tokens or bookplates.
There is still an enormous amount to be done but, you know,
a large part of the collection is out there.
Another thing -- I'm sorry --
I want to say is when we analyze survey results we have
to be careful.
I didn't quite agree with your explanation to why
so many people look at prints.
They look at prints not because of the nature
of the print collection but because
of the nature the print records.
The print records are extremely good.
The prints and drawings records was started out by tutorial use
and trained catalogers and our staff.
And I think also they were the first records that went out
and the keeper of that department --
thanks to whom Anthony Griffiths [phonetic] we have this online
at all -- he promoted this a lot amongst this community.
So I think analyzing the results of surveys has
to be also taken in the larger sense.
We make an assumption about results
but it's not necessarily the true explanation in every case.
>> Melissa Terras: Right.
Yes. I agree.
I agree wholeheartedly with your last point.
I hope I gave a favor of the fact that we have to be mindful
of the type of responses we're getting and how valid
that this thing is and that's really kind
of information studies thing comes
in about all these responses that you're getting as well.
And I'm also not from the museum so I don't have the background
on a lot of these things and I only had 40 minutes
to talk about this today.
There are other things I could talk about in more detail.
I wanted to give a favor of this physical relationship
of museum verses the website.
So, yes, my conclusion was maybe a little bit simplified
and there are certainly more things we can talk about here,
but I hope that I was respectful about the museum as well.
You know, it's a phenomenal investment of time
and effort to do this.
And I know you picked up on all the negative things
that you don't agree with, but I think the fact that most
of the responses from us were showing how well-loved the
British Museum online collection is and that's something you have
to take away from this.
We have got new evidence is to show how well-loved
and well-respected and well-used it actually is.
>> Matthew: Any other questions?
Oh, yes?
[ Silence ]
>> I'm actually wondering how the interface
of the collection looks like.
So is it possible to -- if I found one object,
does it then show like Amazon?
Other people who have looked at this also look at this
or are there links between the collection?
I think it would be interesting to look at.
>> Matthew: I can't show it to you right now, but yeah,
we do have on the page -- so if you found the Rosetta Stone
and that stone, you know that whole -- a stone --
a sculpture and it is from Greece.
It is made of marble and it's, I don't know --
but you can click on the fields where it says "Greece"
and "marble", you click on those terms and you will go
through to the results of all of the objects
that are from that term.
So yes, you can navigate between sets of objects.
There isn't a more direct and inference of where some people
who saw this also saw that but we haven't got
that sophisticated thinking at the moment.
>> And did you do an analysis of that?
>> Matthew: We haven't --
>> Who goes where?
>> Matthew: No, we haven't -- no, we haven't done an analysis
of how many people use those.
>> Melissa Terras: Another thing to say on that point as well is
that the tools that we've got now, two years later,
to do analytics are so much better
than the tools they were two years ago
so you can do much more detailed work flow through sites.
Where people are coming in, where people are going out
and look at the traffic in a lot more detail so that
when we do hopefully repeat this
over a longer period we'll get much more high level results
of people -- that we can actually see these kinds
of things.
>> I had a question which is you talked about the release
of pictures in the future you could do more about looking
into how people reuse the images
that they've found on a collection.
Can you talk about that?
>> Melissa Terras: Yeah,
it's one of the things that's very difficult to know.
If someone's got an image on a web page and done with it
for assay a cover -- there's a system here that you have
to get permission, so we could get access
to the permissions that people have.
But then you have whole illegal use thing, you know,
people are just saving it and using it.
And that's quite interesting because in the last few years,
or last six months or so, you've been able
to actually start tracking and tracing
where these images actually go
to if people are reusing them online.
And they're images that can actually be done to do that.
So over the summer, actually, we're working
with the National Gallery to look
at where their image collections --
their exact images when they put them online --
where those images -- those specific files go to.
Is anyone else using them on webpages?
Where -- what's the distribution network of the images?
And we'll correlate that to the statistics
of the most popular patents
to see whether there's a correlation
between popular access steps and how many times.
But there's always going to be people that save them
to their hard drives and use them --
I don't credit all the time where I get all my images from
and I pretty much -- I try to --
I try to, especially in public lectures like this.
But we've all done it, right?
Right click.
So it's always going to be a bit of an issue.
And I'm fascinated by images and how people love images
and how people use images and it's always going
to be a huge issue especially if we're trying
to articulate the spread of collections and the use
of collections; trying to find a way
to analyze what people are doing with the stuff.
And one of the things we want to follow up on is what people do
with the information and the image once they've accessed it.
And that is the missing link in digitisation just now.
What happens next?
>> Right, thanks.
>> Matthew: There's a question at the front here [inaudible].
>> I'm interested in the failed searches.
If I was in retail, the people that'd interest me are the
people who walk out without buying anything, And things --
why people download images is fine and lot of it's
for academic citation or publication.
But why people fail in what they were looking
for in the first place.
I'll give you a personal example.
I went under the site months ago looking
for the Staffordshire Hoard.
Now that's because as a human being
in the news media I'm very interested in current things
and where objects are and what they are.
And I was fairly disappointed, to be honest, at the level
of information obtained and went away sort of vaguely ungruntled,
We actually went to see the collection in Birmingham
and found it was in New York,
for example, and things like that.
It would be, I think, probably going
to change your curatorial practice over time
if you get the search
which tells you what people were actually looking
for that wasn't yet there.
>> Melissa Terras: Do you want to take
that one because [inaudible]?
>> Matthew: Yes, it is interesting and you're point
about -- and I'll kind of make a larger point as well as is
that the collection online database started off
as a curatorial intern database to kind of manage
and store the curator's knowledge of the collection
for -- And when it was originally written it was never
kind of conceived that it would be read by the public
and it's interesting because you have to kind of --
if the curator's too focused on the fact that it's been read
by millions of people around the world, we can't let
that influence how they use it and how they describe things
and they've got to get on with best way they see fit.
But I do take your point in that we should make sure there's more
information about Staffordshire Hoard even though the object is
not obtained by the museum but we have done we did
and we've done lots of conservation
on it, so, I'll remedy that.
>> I reacted quite strongly when you talked
about images being illegally shown on other websites
and being used by other people and even by right clicking it
for lectures as if that was something we'd wanted
to discourage.
And --
>> Melissa: Terras: Oh no,
I just think it's something that's done and we have
to figure out what happens or how people are using this stuff
to see the dissemination of information.
I don't --
>> I mean it's something surely we want to encourage.
We want people to know what we've got in the museum
and if other people are prepared to do us a favor of showing them
in their lectures and it's something that really encourage.
>> Melissa Terras: Oh absolutely.
You know the Museum of Modern Art in New York;
it's got the most famous art collection in the world, right,
modern art collection in the world.
The reason it's got the most famous modern art collection
in the world is right from the 1960's when it opened --
or 1950's, it allowed people
to use the images wherever they wanted in print media
so they put the paintings in the museum,
became the most famous modern art paintings
because they relaxed the copyright laws.
That's the story.
And that was the U.S. Government that decided to do that;
to take the art world away from Paris and move it to New York.
So, you know, in the operating world,
there is a model for this.
And I'll -- absolutely, people should be used.
What can you do with a little resolution jpeg?
You can put it in a PowerPoint.
You can't put it on a tea towel and sell it.
And that's where the problem is.
The problem is when people began to get money from getting
into high resolution images.
But I think, yeah, low resolution images --
people should be allowed.
It becomes an issue when there's money involved;
when people are creating products.
But it's interesting to see where these images --
So I'm sorry if I gave the impression
that it's the wrong thing to do.
I think that people should say where they got the image from
but that's just respectful, right?
It's respect you take something and it's respectful to say
where you've taken it from.
But I wouldn't want to discourage people
from that sharing that type of stuff.
>> Matthew: In the middle of the [inaudible].
[ Silence ]
>> Hello, I think this is for Matthew, really.
You were saying that you're redesigning the search
at the moment.
Are there plans to actually make that more accessible
to the groups that aren't using it so much and grow
that audience at the moment or are you mainly concentrating
on improving the field of the search
for the academic audience?
>> Matthew: We're focusing on improving the field of search
for everybody so that's easier to search.
One thing we're working on as Tanya kind of alluded
to was reduce the number of failed searches
by having an [inaudible].
So if you start typing into the search field, it will start
to guess what words you might be searching for based
on the words we have in the database.
So which kind of removes misspellings;
it also finds alternative spellings of words
because we've also got those in the database.
So it's going to hopefully improve it
for absolutely everybody.
The other thing is it's not just the search into it;
we're improving the whole of it.
And since we went online in 2007,
we've redesigned the website to be wider, you know,
everyone has wider screens.
The thumbnails used to be quite small.
We're making the thumbnails bigger so when you get
to the search results you can actually see which object --
see the images better so you know which one to choose,
so lots of things like that.
I mean, of course, we're constrained
by the way the database is structured and,
you know, the terminology.
We can't make the terminology simpler, as I said.
We're not going to change the way the objects are cataloged
because it's accessible to the public.
You know, there are two million objects; we can't go in there
and say we describe them all for a different audience.
That's not the idea.
It's kind of whoever wants to use it we'll make it
as simple as possible for them.
Does that answer your question.
There's a question right at the back.
>> Hi. I'm very glad to hear you're making the thumbnails
in the search results larger cause
that would be my number one criticism
of the current British Museum collection is
that the thumbnails are so small you can't see what they are even
if you know what the object looks like.
What I was going to bring up was just to follow on these points
about reuse of images from collection websites.
A little case that you might be interested in is where I work
which is a tape website,
we recently redesigned our collection pages and as part
of that we increased the size of the images
that we've made available online from a previous maximum size
of 512 pixels, which approximately an eighth
of a computer screen area to 1536 pixels,
which basically would fill most of the computer screen normally.
So they're much higher resolution.
And there were a bit of [inaudible]
about this internally about were we giving away all this
wonderful digital content
that our image library could sell the rights to otherwise.
But one of the other things that we did was we add a link
under each image saying "license this image"
which took you straight to the image library.
I think the British Museum has something quite similar
to that on the site already.
And as a result, we've actually doubled the revenue
of the image library since launching that collection.
So we haven't actually --
maybe more people are taking our images for free,
but also more people are buying them as well
as a result of redesign.
So it's not all bad.
>> Melissa Terras: That's good.
Thank you.
>> Matthew: Thanks.
Vicky [assumed spelling].
I'm sorry this is our last question for time here.
>> Vicky: For the people
who don't know what they're looking for,
there is evidence that, you know, they don't know what
to type into a search box
because they don't know what you have
which is perhaps why they start out with something obvious
like Rosetta Stone, because at least, they know you have it.
But there is some evidence that browsing is also one way
of getting around that problem of letting people see.
It's like going to a department store; you get to browse around.
You don't have to go to the department store
and say "Do you have X?"
So are you thinking about a browse capability as well?
>> Matthew: We are going to have on the initial page --
we're going to have links to [inaudible]
but also collections, so kind
of premade searches which we can change.
If there's an exhibition on, we can always kind --
there's a Chinese Porcelain Exhibition we could do an image
with a link to all the results
for China's porcelain and things like that.
So we will do that, yeah, that's planned.
So it's turning into a different.
>> Melissa Terras: No, no, no, it's fine.
This is fine.
I'm, again, very interested to hear all this because,
obviously, I'm not from the British Museum.
I interpret the users but to hear their preferences.
>> Matthew: Okay I think that's all the time we have.
Thank you very much.
I'd just like to ask you to share your appreciation
for Ms. Terras once again.
>> Melissa Terras: Thank you very much.