Introduction
The Royal British Columbia Museum
The Royal British Columbia Museum is a natural and cultural historical museum
and archive in Victoria, British Columbia, Canada. It houses over 7 million
artifacts (Royal British Columbia Museum,
2015).
Selection
I chose the Royal British Columbia Museum [RBCM] Transcribe project because it gave me the
opportunity to transcribe personal journals. I had done this type of
transcription before as part of the Helen Keller archive, but that project had ended
before I started work on this assignment. So the RBCM it was!
I selected the
personal journal of Martha Douglas Harris (1854-1922), the youngest daughter of
Sir James Douglas, first governor of British Columbia. This journal, kept by
Martha Douglas from 1872-1873, covers her trip from her home in Victoria,
through the United States and to England and continental Europe (“Royal
BC Museum | Transcribe,” n.d.)
Interface and Usability
The first thing a user sees is a series of thumbnails for
various collections of documents, with an image and title giving an idea of
what sort of documents the collection contained. This encourages users to
choose the type of documents they are most interested in working with, which I
imagine makes the process more enjoyable and could cause users to remain on the
site longer.
The interface allowed users to see a thumbnail of the pages
prior to selecting one to work on, which permitted me to choose something I had
adequate time to finish in one sitting. The interface was fairly typical,
consisting of a document window with arrow keys for moving around the page and
a magnifying glass icon for zooming in on the image. I found that I was able to
navigate these pretty easily, though the site response was slow enough that it was hard to determine
when I had moved or zoomed enough, because the page would catch up several
seconds after I stopped. The next section of the page was for entering my
transcription, which consisted of nothing but a big white box.
Here's what I entered:
Receipts of Money1872 $ ctsAug 11 Received from Galen 75 ." 18 [struck] Recieved from Bassenger 144 00" 28 Bassenger Ja wth Received f B.H.A $100 00" " Received from GR Patton $5 00Sept 17 Received from [Struck] L S "Joseph the sum of 50 "281441005------2775------282
I clicked a “?” icon, which led me to a page describing
WIKI-style markup, so that gave me the idea that I should attempt to format my
work using that markup. It was very useful for mimicking strikeouts on the
original page, but more complex formatting options from the help page did not
seem to render as expected. Not only are the data items not aligned at all, but
the dashes that should create a horizontal rule created text boxes around some
numbers, instead.
This was frustrating, since the page I was transcribing was
not formatted like a journal, with sentences and paragraphs. I needed to be
able to use columns, and even depict an arithmetic problem. The current
implementation of transcription formatting just does not handle table or column
organization at all. This problem is likely to have affected only a few pages
in the journal, but I imagine other types of historical documents might contain
many tables, calculations, or other non-block formats. For those instances, the
transcription experience are both confusing and frustrating.
The project could
benefit from the addition of more scaffolding, which librarian and blogger
Trevor Owens asserts “allows interested amateurs to participate…through their actions
on the website, without needing the skills and background of a professional” (Owens, 2013). This would probably look
like creating a new web form which broke the task of transcription down into smaller
chunks, in order to take formatting worries away. It probably wouldn’t be
worthwhile for the small number of pages in this database that are tabular in
nature, but the idea is very important to think about when planning a
crowdsourced project. Owens puts it this way: “What expertise can we embed
inside the design of our tools to magnify user efforts? How can our tools put a
potential user in exactly the right position, with the right knowledge, just at
the moment he or she needs it, to accomplish a given activity”(Owens, 2013).
There are no points, no head-to-head competition between
transcribers, and no other overt "game" elements in the interface.
Data and Analysis
One interesting thing for me about this site was a total
lack of apparent tracking data. No login was required, so users could not be
tracked that way to determine how long they stayed on the site or what level of
contribution they made to transcribing. Perhaps this tracking is being done
behind the scenes using IP addresses, but no amount of searching their site,
the RMBC blog, or scholarly sources turned up any tracking or mention of how
the project was going. It launched in 2015, so it's possible that the figures
just haven't been released yet.
There is no publicly available evidence that RBCM plans to
make the dataset public; this doesn't really surprise me, since the tradition
in the museum world is not so free and open as that of the library world. It
will be interesting to see what they do release, since the project is reaching
its anniversary right now. After the grant year has finished and the project
either closes or becomes supported through other means, it seems likely that
more information will come to light, either as formal published work, or in the
blog and other publications released by the museum.
Goals and Sustainability
The goal of the project is very simple: "to improve the
Royal BC Museum and Archives’ public accessibility by turning handwritten,
audio, and video records into searchable data."(“Royal
BC Museum | Transcribe,” n.d.) On first glance it seems that
the plan consists solely of leveraging volunteers for the purpose of
transcribing the materials faster than could be accomplished in-house. It seems
unlikely to be the only real goal, so future reports may reveal more.
NYPL Labs: What's on the Menu?
The What'sOn The Menu? project by NYPL Labs extends an earlier project to digitize the now 45,0000-item collection of menus
begun by library volunteer and schoolteacher Miss Frank E. Buttolph (1850-1924).
This digitization project asked users to transcribe and verify the contents of
many vintage and historical menus from all around the city and other places,
too.
Selection
I selected it because
in the area of user engagement and interaction it is very like RBCM's
Transcribe, while it is very different in transparency, data availability and
maturity.
Interface and Usability
The interface was very similar to the RMBC site, but the
experience of working with it was more pleasant because the server was more
responsive. Another issue that made
transcribing easier for this project is the fact that menus are typeset rather
than handwritten; I am getting better at reading historical handwritten
documents, but it's still a challenge compared to printed text.
Because the approach of this project is more granular in
nature, asking for transcription of individual dishes rather than entire menus,
the process of transcription is much less frustrating. I don't need to attempt
to format the transcription at all, so I'm free to focus on the content. I
recognize this now, having read Owens, as a “scaffolded” interface (Owens, 2013).
The menu project is much older than the RMBC one, so many of
the menus have been transcribed at least partially. Most of the work with which
users are initially presented consists of verifying existing transcriptions, or
pages that have very little text to transcribe. Perhaps verification is
presented first because it is somewhat easier to do than transcription,
allowing new users to ease into the process. The web page design here is more
retro-styled, and uses cute old-fashioned clip-art hands to point out the best
place to begin.
Again there is no overt gamification in the project.
Transcribers work with the menus because they are interesting or because they
want to help the library, and not for an
artificial competition or reward.
Data and Analysis
Because this is a more mature project, there is information
regarding usage numbers and the overall sense of progress. I was surprised to
discover that there are "approximately 45,000 items" in the physical
menu collection. Clearly there is a lot of transcription still to come! The top
of the project's home page shows "1,332,229 dishes transcribed from 17,545
menus" to date, making it very clear to every visitor that the collection
is big and that a lot of work has already been done(“Whats
on the menu?,” n.d.).
The data set which has been generated by the
"crowd" is also open for download and use by the public, in CSV
format for use with spreadsheets, or as an application programming interface [API]
for developers who may wish to use the data set from within their own projects(“Whats
on the menu? Data,” n.d.). This is a very sophisticated approach that web
developers will be comfortable working with, positioning the project to get
outside the library even more than it already has.
The spreadsheet data focuses exclusively on the menus
themselves; no user data is present. This shows a focus on the outcomes — more
digitized and verified menus — instead of who is using the project and how much
time they are spending. True to the bragging on the project’s home page, the
spreadsheet data comprises 4 separate tables and includes almost 1.5 million
individual menu items, from over 17,500 menus.
Goals and Sustainability
The project launched in 2011 with grant support, and still
proceeds today, demonstrating strong sustainability. I wonder if this is a
function of being part of a large and well-funded library system, or the press
attention the project received, or simply the fact that people find the menus
entertaining and useful. The WOTM about page speaks of the collection being
used by chefs, historians, and novelists in their research (
n.d.).
Conclusion
Each of these transcription projects is a classic case of
not needing to "gamify" the process, because the documents themselves
engaged the volunteers. Therefore, there are no "points" associated
with it, and no explicit rewards for completing transcriptions. I have played
with other projects that do use things like points and levels to support user
motivation, and I imagine that might increase the length of time a user spends
there, especially people who really enjoy gaming.
References
Owens, T. (2013). Digital
Cultural Heritage and the Crowd. Curator: The Museum Journal, 56(1),
121–130. http://doi.org/10.1111/cura.12012
Royal
BC Museum | Transcribe. (n.d.). Retrieved April 12, 2016, from
http://transcribe.royalbcmuseum.bc.ca
Royal
British Columbia Museum. (2015). Royal BC Museum Corporation Statement of
Financial Information for the Year Ended March 31st, 2015. Retrieved from
http://royalbcmuseum.bc.ca/assets/CSCD-RBCM-2014-15-FINAL-MINISTER-APPROVED-ASPR.pdf
Whats
on the menu? About Page. (n.d.). Retrieved April 12, 2016, from
http://menus.nypl.org/about
Whats
on the menu? Data. (n.d.). Retrieved April 12, 2016, from
http://menus.nypl.org/data
Whats on the menu? (n.d.). Retrieved April 12, 2016, from http://menus.nypl.org/
No comments:
Post a Comment