Help improve Transcribe Bentham
By Tim Causer, on 1 October 2012
From today—as part of a project entitled the Consolidated Bentham Papers Repository—Transcribe Bentham will be supported for two years by the Andrew W. Mellon Foundation’s Scholarly Communications Programme. A significant part of this work will be to make modifications and improvements to the transcription interface, in order to make the transcription process more straightforward for volunteers. The code for the updated Transcription Desk software will, like its first iteration, be made available on an open-source basis for others to re-use and customise to meet their own needs.
Through a survey carried last year (the results of which will soon be published in Digital Humanities Quarterly), volunteers have already given us a few ideas for alterations which they would like to see made. These include:
- Changes to the way the markup is added. Responses to the survey indicate that adding of XML tags was a significant issue, and may have dissuaded some from participating. though the Transcription Toolbar was designed to make adding markup as easy as possible, markup was regarded by more than a few respondents as an extra complication when trying to decipher Bentham’s handwriting (which is, after all, the purpose of Transcribe Bentham). One solution is to introduce, as an alternative, a What-You-See-Is-What-You-Get interface, so that transcribing will be like typing in a word-processor. In this scenario, the transcription toolbar would be done away with, and transcribers would not have to concern themselves with visible markup at all. Volunteers would—just by way of example, these are not set in stone—indicate a line break simply by hitting return, a paragraph by pressing return twice, and indicate underlinings or deletions by selecting the portion of text, pressing a button, and the underlining/deletion would be rendered while typing. The XML tags would thus remain behind the scenes and would not clutter the transcript.
- A more flexible image viewer. The current interface, where the image is alongside the text box where the transcript is entered, was also seen as problematic. We will aim to introduce a more flexible image viewer (image box above text box? A floating image window? a resizable image window?). A floating transcription toolbar would also prevent constant scrolling up and down.
- Distinguishing between locked and fresh material. It is also clear from volunteer responses that we need to make it easier to distinguish between complete and locked, partially transcribed, and untranscribed manuscripts. While we have introduced lists (e.g. the untranscribed manuscript list) and they have their uses, these have to be updated manually and—owing to human error—are not always fully accurate. We will look to introduce some form of automated system to tell the different types of manuscript apart (some form of colour coding?)
Sadly, we can’t do anything about the state of Bentham’s handwriting, and/or composition! Changes which would be made are all about making life easier for transcribers, so this will be very much evolution rather than revolution. Volunteers will be invited to test out any alterations made to the interface.
These are just a few of the ideas we have gleaned from volunteer suggestions, but we are keen to hear more (crowdsourcing inspiration, if you will). We would love to hear from interested onlookers, as well as those who have had hands-on experience with the transcription interface. What alterations would you like to see made to the Transcription Desk, or would you recommend be made? If you are a volunteer transcriber, what changes would be most beneficial to you?
We would like to hear any ideas which you might have. You can leave comments here on the blog, in the Transcription Desk’s discussion forum, or we would be delighted to hear from you via email. Alternatively, you can leave your suggestions anonymously on this survey.
We look forward to hearing from you, and do let us know if you have any questions!
10 Responses to “Help improve Transcribe Bentham”
Tim Causer wrote on 3 October 2012:
Thanks very much for these suggestions – just the sort of thing we’re after. The autosave is an excellent idea, as is the disabling of certain keystrokes (I’ve lost the number of times I’ve messed up the focus of my manuscript by accidentally flicking the mouse wheel).
I’ll look into de-cluttering the front page, and good points too about the discussion forum – that could probably do with a revamp.
Your experience of the XML certainly mirrors many of the responses we had to the user survey, in that the XML is more than annoyance. While some have taken to it well (or perhaps just put up with it), we’re pretty sure that it’s put a lot of people off participating in TB. The easier we make it – and we hope the WYSIWYG interface will remove the need for transcribers to have to deal with visible XML, the more straightforward participation should be (we hope).
thanks again, and we look forward to further discussions over this in the coming months!
all the best,
Peter Hollis wrote on 3 October 2012:
All these are good suggestions, and I was also going to suggest autosave, as it is so annoying to have the connection go down and lose a whole chunk of work.
I like the idea of having markup hidden in the background, so long as it would be able to cope with the many complex additions within additions within deletions plus underlinings etc etc. These are the bits where so much care is needed to get it exactly right, and even then it can look a mess.
The pages where there are charts, tables, flow diagrams etc are virtually impossible to transcribe coherently so any method that can cope with that would be valuable. Similarly, it can be very difficult to discern whereabouts in the text to insert marginal notes, sometimes the order of paragraphs and sections is not clear, and deciding whether a chunk of text in the margin is really a new section or paragraph. So a format which enabled the transcription to at least roughly mimic the layout of the page could help.
I hope this is of some use. I’m sure some of us would be happy to test out new software before you go live with it.
Peter Hollis wrote on 3 October 2012:
Also, a way of distinguishing a main heading from a sub-heading or
Tim Causer wrote on 5 October 2012:
Thanks very much for your suggestions too – always good to hear what users are finding to be the most troublesome issues with the transcription interface.
The XML is clearly something we are going to have to look at. I know what you mean – when I’ve transcribed something with a few additions it can be extremely complicated (not to mention time-consuming) to find any errors I may have made in the markup, and to distinguish between the XML and the transcript text itself.
You’re quite right about charts and tables too: JB uses them fairly often, so we could do with some standardised form of notation for these. Ditto on placing of marginal notes.
Thanks again for these ideas – the project team will be meeting soon, and these suggestions will be discussed when we talk about changes to be made to the interface. Anything we can do to make the process more straightforward will be looked at! And we will indeed be coming to you and other regular transcribers when it comes to test-driving the new interface, to see what works (and what doesn’t!). We’ll keep you all informed in the meantime.
best wishes and many thanks,
JFoxe wrote on 8 October 2012:
The suggestions from Lea and Peter are excellent. One more suggestion occurs to me, although it only applies in a few cases. If it would be possible to rotate the text, it would make the occasional sideways text easier to decipher, not to mention the few cases where a part of the text appears upside down. I can manage lying on my side, but I’m not so good standing on my head!
Tim Causer wrote on 11 October 2012:
Many thanks for this, and you are quite right about the importance of being able to rotate the image – finding a solution to this is certainly on the ‘to do’ list. Transcribing Bentham is often painful enough without us causing everyone cricks in the neck!
best wishes and many thanks,
Transcribe Bentham » Blog Archive » Progress update, 6 to 12 October 2012 wrote on 12 October 2012:
[…] again to all those who have made suggestions on how to modify and improve the Transcription Desk in order to make the transcription process simpler. We have had some great ideas, and look forward […]
LMason wrote on 23 October 2012:
It should be possible to do almost everything you want with XSLT/HMTL? I haven’t looked at v5 yet, but the only problematic issue for v4 would be Joy’s suggestion and that can be resolved by printing the text to a JPG and rotating the JPG for display. All of this is achievable within a relatively simple WYSIWYG interface.
XML, while not ideal for display to non-techies, is the wisest storage medium, but real-time translation would keep both editors and readers happy.
Transcribe Bentham FAQ | UCL Transcribe Bentham wrote on 27 March 2013:
[…] This has been a common issue,, and an improved manuscript viewer is a priority in the ongoing modifications to the transcription platform. In the meantime, there is a workaround. If you right click in the white space below the Zoomify […]
The improvements you mention all sound worthwhile. A couple that I’d find very helpful as well: (1) automatically save transcriptions as they are created (every X seconds, as Word does). That way, when you’re knocked out of the system, as I have been, not much work is lost. (2) Locking the keyboard while transcribing so that accidental keystrokes don’t knock you out of the system (see 1).
As for the Transcription Desk, the initial screen seems a bit crowded. I suppose the names of sponsoring organizations need to be up front with all the disclaimers etc., but all that is information that doesn’t serve to lead one to keep reading. And what a shame the Discussion Forum is so little used. I’d find it more inviting if the year of the previous responses were shown; more relevant than time of day, or even day of the week, I think. I’m also not sure why the Forum is broken into subject areas. That might make it easier to find earlier posts, but figuring out what belongs where is just an extra nuisance, and the divisions take up more room on the screen as it is.
Finally, the XML system is truly irritating, but I don’t imagine much can be done about it. It’s Incomplete (no way to show position on the page, as in JB’s charts), it’s Sloppy (those extra s showing up when there are two additions above the line), and it’s Inflexible (handling some of JB’s weirder bottom-line effusions). Otherwise it’s fine.
Thanks for your patience with all this. I appreciate your asking, and I look forward to hearing/seeing what others think and what changes are made.