memoQ and Team Project 4

And finally onto memoQ[1], the CAT tool we used for the final Leeds MAATS Team Project this year. I intend to start with a couple of quick points before getting stuck into the nitty gritty of the pros and cons of memoQ. I will then proceed onto a more detailed account of the project itself.

First and foremost, there are a few features of memoQ I know about but have not yet got round to trying – they could be pretty useful (if they work). The first is the concordance dialog, much more inclusive than other CAT tools as it allows you to include not just the TM but also Live Docs corpus (which you can build up yourself with mono-/bilingual documents, or alignment). The built in web search could make for increased efficiency, too, allowing the user to add their preferred tools.

There are a couple of memoQ features I was not so keen on, however, primarily some processes being unnecessarily long. To set penalties, for example, the user must make a copy of the default settings and edit the copy, rather than just altering the settings. Meanwhile, to carry out a pseudo translation they have to go to the settings for MT, uncheck everything except pseudo translation and then pre-translate their chosen documents.

My final issue was with there being no button to (de)select all people when creating a “Handoff” (more information on Handoffs to come). It’s a minor flaw, but when dealing with a huge document, divided between a large number of translators the user might want to create just a couple of the handoffs. To have such an option would make a really useful update to the software.

Overall, the Handoff packages are a great idea, and set up so that they can be assigned to specific people. The options also exist to (not) include skeletons[2], or (not) to join or split segments. In most cases, a PM is unlikely to want to give the translator either of those options, but you never know. The Handoff files created are automatically labelled with the name of the translator/reviewer the document is assigned to – both cute and useful – whilst the file extension also changes, to show whether it is the “out” or “back” file, each of which has its own clearly identifiable icon. The Handoffs are automatically stored in the project folder – logical and helpful.

Another useful feature that works neatly alongside Handoffs is the ability to create “Views”. These allow you to combine smaller documents, or split larger ones: better than Trados, because you can split larger ones, but not quite so efficient as DVX3 as the process is manual – the user enters the numbers of segments they wish the View to include. Whilst this can result in more logical division points than DVX3’s automated process, it certainly means a bit of trial and error and therefore a lengthier process. Views can be assigned to translators, and therefore Handoffs created that incorporate Views.

As you can see using Handoffs and Views can really help PMs but, on top of the lengthiness of file splitting, there are still drawbacks. After reimporting the .mqback files, between the translation and revision stages of our project, we wanted to create the review Handoffs. However, the new Handoffs could not be opened as a new project by the reviewers, generating an error message that suggested memoQ wanted to import the file as an update to an existing package [3]. Obviously, this is useless: the whole point is the Reviewer is being sent something new to look at; furthermore, we did not have this issue when sending the original Handoffs to the Translator. The solution was to wait until the Handoffs had been returned and delete the Handoff history but this creates three issues: firstly, it takes longer; secondly, it posed issues with missed deadlines (translators are human, anything can happen!); thirdly, what if you want the entire Handoff history (showing when the files were sent out or delivered at each stage) at the end of the project?

The exchange of files between memoQ and Trados via a TRADOS-compatible RTF is another valuable memoQ functionality, foreseeing the probability that not all translators use the same software (although not foreseeing that they might use something other than memoQ or Trados – I am sure there are reasons for Trados being the CAT tool that memoQ works with). However, when updating the memoQ project from the translated RTF there are some tag issues: the tags were returned differently, and memoQ refuses to import any segment containing different tags to the source segment. Unchecking the option, “Do not import update if source segment is different” appears to be a quick solution to this; however when it comes to generating the target file, memoQ will not do so until the tags are identical to the originals. I am aware there may be a solution to this, but I am not yet sure at what stage – perhaps there are options to select when importing the source files into memoQ that would have avoided this issue.

What memoQ does do at this stage is collate all the errors on one page, where they can be corrected. It also automatically updates segments with the same sets of tags as you confirm changes. Regardless, it still makes for a long-winded and somewhat tedious process (although I must point out here that I do really like memoQ’s tag insertion functionality!).

Overall, I was seriously impressed with the range of options available to a memoQ user, particularly from a PM point of view. I do feel adjustments could be made to the software to make certain processes a little more efficient (the changing of settings seems so complicated in so many ways, compared to other interfaces). It does seem to offer a lot of things other CAT tools do not, though, and I am looking forward to hearing responses concerning some of the features I have not yet tried, or regarding solutions to the issues I came across and have written about here.

*          *          *

And now – Team Project 4. Taking place in 3 locales, operating between 5 CAT tools and involving 77 participants and 5 disciplines, this project is the most complicated ever undertaken on the CAT tools module at Leeds. I am certainly proud to say that I not only took part but took on a leading role, as one of 10 PMs. The University of Leeds collaborated with the University of Texas at Arlington, US and the Universa v Llubljani (University of Ljubljana), Slovenia, to translate, proofread and subtitle materials in 13 language pairs; we even commissioned some DTP work and consultancy roles.

I will not go into too much detail on how projects work on the MAATS course at Leeds, having done so before, but definitely have a read of my blog on a previous project (using DVX3 and the TeamSERVER) to gain an insight.

Having so many PMs, we initially wondered if this might pose a challenge to working as a team: taking on the three-week project in the midst of exams in the first week, and thereafter alongside our usual lectures, posed difficulties to being in the same place in the same time. However, there is no need to be modest about how strong the teamwork was. Starting with an initial meeting, the day after our project briefing with the “client”, we delegated tasks (e.g. researching rates, drafting emails, creating folders on Google Drive, etc.). We also split the language pairs between us, which we each followed throughout the project: for each language pair, a PM had to examine them before sending them out, divide them up using memoQ, receive the return file, then generate the target files and carry out necessary post-editing. Ultimately we did spend a lot of time on this project, arguably decreasing its cost effectiveness; however, dividing up tasks limited time spent and it must also be said that we were required to submit our own deliverables and instruct the “freelancers” concerning theirs, for the sake of assessment – a ‘real-life’ project might not involve such extensive file preparation.

Predicted timeline and Final timeline for TP4, made by fellow MAATS student Catriona Burns

It must also be mentioned that we kept each other going by bringing in treats: the steady flow of biscuits, brownies, bananas and flapjacks in this project was second to none… perhaps another bonus of having such a large PM team!

Taking example from previous projects, we also decided to build and maintain a corporate image throughout the project. The forum we created (for us to discuss without endless, back and forth, confusing email threads and for the freelancers to discuss and share terminology where we could keep an eye on it), our email signatures, our instruction PDFs and our quotes and invoices all carried the same green and orange colour combination. We used templates so that the creation of documents was only done once (and adjusted by each PM).

So what did we do well? The ‘clients’ told us there was not much they could say against what we did: our deliverables were excellent, they liked the use of templates and signatures, teamwork was great, we ‘educated the client’ well (especially with regards to Texas and Slovenia, who were working in different setups). They also praised our hiring of consultants (namely for DTP on the EN-AR files).

However, nothing is ever perfect! We were told PMs almost never ask a freelancer to quote blind, and that we should ideally send whole/part of the file before asking for a quote (we sent a brief explanation of the subject matter and a rough word count). Furthermore, the quotes were a little too detailed – we were reminded the company contacting you may not be knowledgeable about translation, in which case extensive detail on rates for fuzzy matching does not need to be included.

Personally, I attempted to really speed up some of the processes, using Google Sheets to create my quote and Mail Merge to change names, language pairs and the details in emails and on quotes. Unfortunately, there were areas where a little more checking over this would have lead to me making less stupid and embarrassing mistakes. For example, I changed the formatting of Sheets so that the number of decimal places was only two, but did not realise the formulae took into account the original decimal. In some cases, Sheets simply did not update automatically. Ultimately, the result was a highly inaccurate quote.

So that was TP4. It was a huge amount of work but well worth it. As much as I was relieved to have a break at the end, to work so hard for the desired outcome and for such positive feedback was massively rewarding, certainly allowing me to see myself taking on such a role as a career in the future.

Perhaps with less biscuits.

*          *          *

[1] We used the translator pro and project manager versions.

[2] The skeleton allows the translator to generate the target file after translation.

[3] “This package contains an update, but memoQ cannot find a suitable project on your computer. Please make sure you import the original handoff package, then proceed with the import of the current update package.”


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s