Kilgray TM repository: a New Home for Translation Memories

As Kilgray Technologies made memoQ 5.0 Release Candidate available for download right on queue last week, there is another piece of Kilgray news I’d like to share with you.

The lesser-known but aptly named TM repository was launched recently by the makers of memoQ and offers an interesting and fresh approach to Translation Memory server products. This application apparently pre-dates memoQ but wasn’t launched commercially until this year. Since then, Kilgray have been gathering early adopters feedback, which they are planning to include in a version 2 sometime next year.

TM repository is made-up of 3 components: the database, the business logic and the web-based interface. It is built on SQL technology and comes in 2 editions depending on the number of users required.

The idea behind any TM server product is to provide a central location where all users in a supply chain can access the same and latest version of Translation Memories. Different Localisation Managers have different TM Strategies which are often dependent on the CAT tools or even the version of the CAT Tools in use by the Assets owner and their LSPs. Important choices have to be made in terms of Maintenance, most of which have to do with how best to archive TMs for re-use. Working from project-specific TMs only gives smaller leveraging power and little version control ability.  Yet it is sometimes the chosen path, simply because it seems more manageable. On another hand building and maintaining Master TMs containing all segments ever translated, or even chunks of them organised by Product lines, Business Units etc. requires a sustained management effort. For instance, when there are terminology updates a linguist should implement global changes by batch editing Translation Units. They may spend time fixing old Translation Units (TUs) which will never be used again. It may also be difficult to find linguists with the skills to directly edit the TMs for all languages. More often than not, Master TMs which are not integrated with a Translation Management System will contain errors, deprecated terms, duplicate TUs with alternative translations etc. and require clean-up. The Project TMs-only route will always underperform in terms of ability to re-use existing translations and ensure consistency, but the Assets owner are still left to evaluate for themselves which option is the best for them.

TM repository is a solution to a lot of these common problems:

• It enables the Assets owner to create a single Online TM Database containing all TUs, for all projects, and all language pairs.
• The flexible descriptive fields (metadata) allow the TU’s to be tagged precisely.
• This metadata can then be used in Queries for smart filtering during Maintenance or Export
• TMX Imports let users add to the database from virtually any system
• TMX Exports  permit the extraction of Project TMs, which can be reimported after use and update
• Exports can be customised for the CAT tool in use through customisable Mapping. Query results (i.e. Project TMs) will contain metadata compatible with the target translation tool.
• Refined Maintenance is enabled through features such as Search and Replace of text or metadata, or the use of deprecation settings by which older TUs can be hidden from search results.

Related articles

• Mr. Q Brings Change Management to the Localisation Continuum (localizationlocalisation.wordpress.com)

Alchemy Catalyst 9.0: A Practical and Visual Guide

I recently had the welcome surprise of finding an invite to a Catalyst webinar in my Inbox. It was with great anticipation and a touch of nostalgia for my Localisation Engineering days, that I clicked on the link and joined the meeting to discover what Alchemy had been up to.

I soon realised that a practical user’s guide would be the best way to cover this on Localization, Localisation. The Alchemy Software Development website already lists What’s New in this release so rather than analysing the differences between Catalyst 8, for which we did a complete Launch coverage and Catalyst 9, I’ve put together a step by step tour based on the demo.

This article can be used by Localisation Engineers and Translators alike to preview the Catalyst 9 interface using the 30 or so screen shots included (see after the slideshow for full screen versions), and also to read through some recommended processes and tips, adding to my past article on the Leverage and Update Experts.

This slideshow requires JavaScript.

Creating a Project

The User Interface remains the flexible and now very familiar .net window, with its various docked panels and tabs. It’s also a stable interface which will cause little or no navigation headache to even the most novice user.

The first operation when getting started with Catalyst is to create a Project file, or TTK file. This is easily done by using the File – New menu and following the basic steps.

You will notice in the screen shots that the example used includes varied sample files such as compiled help (.chm) not requiring any source or project files, and wpf executable.

Preparing a Project

After the creation of the TTK, source files can be inserted either using the Insert Menu item or a context menu in the Navigator tab. Folder structures can also easily be used.

Once the files have been inserted into the TTK, it is time to prepare it for leveraging. This operation of consists mostly of locking non-translatable strings and sub strings. It can be tedious on a brand new Project but the work done can be completely leveraged to the various language TTKs as well as any future versions of the project.

The lock keywords functionality has been improved in Catalyst 9: the txt file which the project’s keywords list is now automatically generated in the background as soon as the user locks a keyword.

Once a keywords list has been created, it can in turn be used to automatically lock the listed keywords in the remainder of the project.

Another thing to note is that Maximum String Length can now be set on a batch of strings at once.

Leveraging previously translated content

Apart from Leveraging from the TTKs of previous projects, Catalyst supports leveraging from a variety of Translation Memory formats:

• Translation Industry Open Standard (*.tmx)
• SDL Trados 2007 (*.tmw)
• Wordfast Pro (*.txml)
• Tab-delimited (*.txt)
• Alchemy Translation Memory (*.tm)
• Alchemy Catalyst (*.ttk)
• Alchemy Publisher (*.ppf)

Alchemy Translation Memory is a new proprietary format used to create Master TMs from completed TTK projects. This format allows to store Catalyst-specific context information such as the context (Dialog box ID, Menu Item etc.), which can later improve the quality of leveraging by providing Perfect match. In Catalyst terms, a Perfect Match is a 100% match located in the same Dialog, Menu etc).

Alchemy Publisher, Wordfast Pro, Trados 2007 or the nonproprietary TMS are also present provide compatibility with other TM format Catalyst might have to coexist with.

Noticeably, Trados Studio 2009 TMs (.sdltm) still do not appear to be supported.

Batch processing

The process recommended by Alchemy is to create an English to English Master TTK and then to automate its duplication and pre translation for each target language in the Project.

This is an area where Catalyst 9.0 does seem to bring a good bit of novelty:

• With Catalyst 7, engineers had to manually duplicate TTKs.
• Catalyst 8 was a bit more helpful and created Project folders for target languages and project resources.
• In Catalyst 9.0 however, the Job file and Scheduler take care of a lot of the repetitive tasks associated with preparing a new Project.

The Create Job Expert lets you use the Master TTK as a template to create project folder structure and corresponding target language TTKs.

Meanwhile, such tasks can also be added to the Scheduler. This new queuing system allows the user to start working on the next project while it processes queued tasks in the background.

Automation

The Command line automation has been improved since Catalyst 8 to include Analysis. The complete Catalyst localisation process can now be automated.

Catalyst 9.0 Developer Edition also includes the Comm API which lets advanced users script TTK operations all the way down to string level, and output automation reports  in txt or xml format.

Ensuring Quality and Consistency

In addition to Translation Memories, Catalyst 9 also supports several Glossary formats:

• Text files, used in Catalyst since the beginning (.txt)
• Terminology Exchange Open Standard (.tbx)
• Translation Memory Exchange can also be used for Terminology (*.tmx)
• SDL MultiTerm and MultiTerm Server

Validation still takes two forms: the Expert can be run to perform global check, and inline validation can also be switched as a non-intrusive real-time quality control. If a potential error is found, a flag will be raised through the bottom pane, but Translators will not be interrupted. They can simply go back to the issue by clicking on the notification once they are ready to attend to it.

The Thumbnail view seems to be a great tool for engineers regressing bug. It gives a preview of all dialogs in a TTK and lets you click the one which matches for example the screen shot in a QA report and brings you automatically to the location of this dialog in the ttk file.

Translating in Catalyst

The Concordance search and Translator toolbar do not appear to have been changed. Both were introduced with Catalyst 8 where there was strong focus on improving the user experience from the Translator’s point of view, and they seem to have delivered.

The new Re-cycle button is a result of the same ambition. New translations can be propagated to entire project by using the current project as an inline TM in the background. Layouts are not recycled but fuzzies are supported.

Clean up Expert

Finally the Clean up Expert has also receive some improvements. Like for all Experts, it is recommended to close the Project file before running it, and then select the file(s) to process from the Expert’s General tab.

Clean up now creates a postproject.tm Translation Memory and generate supplied assemblies for .net.

Conclusion

in my opinion, this new generation of Catalyst still offers a great solution for visual localisation. Although the differences with Catalyst 8 may not may not make a bullet proof case for immediate upgrade, the 25% discount currently on offer does represent decent value.

SDL TMS 2007 SP4: Some Comments from SDL

Here are some interesting comments from Paul Harrap, Product Manager for TMS at SDL, in reply to my article on SDL TMS 2007 SP4. I’ve also included my own response afterwards.

“(…) I’m very pleased to see we’re getting some coverage in the blogosphere. I’d like to thank you for taking the time to write us up and your positive feedback on the product generally and our new SP4 specifically.

I accept that there’s still some work to do in the product with relation to TM maintenance. As your article accurately reflects, the contents of TMs are updated in TMS in very specific places in the workflow- typically after one or two cycles of review – and what content goes into which TM can be carefully controlled. This is very much by design. We see Translation Memory as the crown-jewel of the linguistic assets of the enterprise customer and so contents are tightly regulated by TMS.

However, we have to acknowledge that bad content can creep into TMs over time – there might be an error in review, or some customers might not review translations quite as thoroughly as others. The changes we made in SP4 to allow the import of files directly into TMs is a response to this requirement. The enterprise can now add/replace contents of a TM directly, without reference to a specific translation job or workflow, as an administrator-level function. This can allow people to quickly and painlessly correct known-bad TUs.

We’re considering including the ability to search through, browse and directly edit the TUs in the TMS browser environment in a future release. While I accept that this is a lacking feature, I wouldn’t concur that we should be putting such power in the hands of the vendor or the freelancer. Seeing the TM as a hugely valuable asset for the enterprise, I expect this is the sort of feature and capability that most enterprises would want to keep in-house.

On the integration with SDL MultiTerm, I very much see a distinction at the moment where TMS is a consumer of Terminology and MultiTerm is the owner of it. Over time we will see much tighter integrations between the SDL products, so the lines between TMS and MultiTerm will very much start to blur, and we have plans to introduce workflow capabilities for term lifecycle management.

On the issue of uploads and downloads and working offline, I think a lot of people would very much agree with you. The single largest corporate user of SDL TMS is… SDL! We have dozens of translation offices around the globe, all of whom deal with the upload and download of files to and from TMS servers based in our hosting centre in London on a daily basis. What tends to drive people offline is the featureset available in the desktop tools. SDL Trados Studio, and its predecessors SDL Trados TagEditor and SDLX, are very powerful productivity tools for the translator. Replicating these features in an online translation environment is a monumental task and it’s  something we are investigating.”

First of all, I would like to thank Paul for this input. Since the ramp up of Trados Studio over a year ago, SDL have made a sustained effort to listen to their user base. The TMS section proves here that they are keeping with this policy.

On the topic of TM Maintenance, which is very close to my heart, I think the business model Paul is presenting is either slightly outdated or, more likely, is missing on a part of their customer base.

From my experience, the outsourcing model has developed so much during this recession that at least in some cases, big enterprises (i.e. the TMS customers) no longer employ Translation Memory management experts. These positions are filled by technicians employed by the LSPs.

Another point is that while TMS customers use Review routinely, they also cannot afford to review all the content they output. Most of the big players have either implemented or are looking into models which allow them to reduce their review cost for languages where the quality is considered stable. This means that TMs may be updated in TMS with contents which hasn’t been reviewed, and consequently that linguists must regularly inspect the TMs and fix any inconsistencies in legacy TUs to prevent reoccurring errors.

I maintain that this task must be assigned to a linguist, and the best placed to do so is a senior Reviewer. Not all of these are in-house, by far.

SDL TMS 2007 Service Pack 4: Love and Hate

I always find it challenging to get a fair idea of what Enterprise tools can do before making a purchase decision. There is so much involved in setting them up that even if a trial version is available, the efforts required to perform meaningful testing are prohibitive.

Many such applications do not come ready out-of-the-box and require extensive customisation before they can be tailored to fit a specific business model.

This is why many purchase decisions are executive decisions, based on ROI reports and presentations showing what the software does. A demo might be setup for you on a dedicated server by the sales person, and you’ll be left thinking “hum…surely it’s not that simple”. This is also why 10 times out of 10, these pieces of software come with a Support package which lets you install regular and much needed updates and bug fixes.

It doesn’t have to be this way!

If you have the opportunity, go knock on a few door and try to find a company nearby which uses the software in a production environment. Contact them, ask to visit, get an independent demo. From my experience (not based on TMS that time) most people will be more than happy to tell you how much effort it took to setup, how many features still don’t work, but also how much their productivity has really increased and perhaps even how many of their employees have done a thesis on the subject! Bottom line: get real-life advice!

SDL TMS, or Translation Management System, is one such behemoth application. Trying to find independent information about TMS on the web is a challenge. In fact, even finding official information can prove frustrating. As for Special Interest Groups… those I found were for customers-only. It seems it’s buy first, we’ll talk later.

So what’s the big deal exactly? Well I’ve been working with TMS 2007 for about a year now and I have a few things to report: some good, some not so good.

What it does well

Let’s start with positive thoughts.

TMS is a workflow tool, designed to connect a customer directly to it localisation vendors and all their armies of sub vendors. It handles big volumes and short turnarounds really well, and is reasonably good at supporting your Translation Memory and Terminology Management needs. It also offers the reporting facilities necessary for all members of your localisation ecosystem to invoice each other, and you.

TMS automates part of the role of the middle men, and is ideal for localisation consumers with a constant stream of translation, especially if they come in the shape of numerous small projects.

Multiple alternative workflows can be set up, depending on vendor selection, TMs to leverage against, TMs to update, need for Linguistic Review etc. Once the correct workflow is selected at the job creation stage, you can be sure it will go through all the steps required. There is little or no human error possible, at least not in scheduling and assigning tasks to the right participant.

TM updates are handled automatically, literally seconds after the last human input in the workflow.

Where it lacks

So are all the vendors really gathering orderly around the assembly line and localising thereafter like a happy family?

Not exactly. There are a few snags.

My main grief is around TM Maintenance or the lack of it. Because TMS automatically updates the Translation Memories at whatever stage of your workflow you told it to, manual editing of the TMs has been neglected. A user can perform a Concordance search, but it is impossible to edit the Translation Units found. One cannot use TMS to fix inherited inconsistencies or any error found in legacy TUs.

This makes implementing Global changes a very untidy task: one needs to connect to the TM Server (hosted by SDL in most cases) using SDLX 2007 Professional. This, to me is total non-sense and here is why:

1. increasingly, the business model in Localisation is outsourcing.
2. once localisation is outsourced to agencies, these subcontract Single Language Vendors, who themselves might only be sub-contracting to freelancers.
3. less and less Localisation consumers employ in-house linguists.
4. their remaining in-country staff is Sales and Marketing, and has much more pressing matters to attend than editing TMs.

Now which version are these freelancers more likely to have? SDLX 2007 Professional (€2,995) or SDLX 2007 Freelance (€760)? I think you probably guessed it. SDL’s licensing model prevents linguists from maintaining TMs in TMS and seemingly forces corporations which bought TMS to support their outsourcing setup, to fix TMs in-house!

There are some workarounds to this, but for a piece of software of this caliber, I think this is a pretty shocking limitation.

The integration with MultiTerm has similar issues: only some of the functionality are available through TMS, the rest including editing Term entries has to be done using MultiTerm Online or Desktop.

Performance issues also tend to drive a lot of linguists offline! Depending on their setup, a lot of them find it more efficient to download jobs, translate offline in SDLX and upload the finished work back into TMS. While there is technically no difference in the end result, this is a disappointing interruption of the workflow.

Service Pack 4: An End to the Suffering?

Squeezing under the gate at the last second, like Bruce Willis in a classic movie, TMS 2007 Service Pack 4 sneaks in before the long-awaited SDL TMS 2010 and comes to the rescue.

With TMS 2010 now possibly slipping into 2011, it is a welcomed addition particularly due to the improvements it brings. Here are the most significant end-user facing features:

Browser support: IE 8 support added (IE 6 removed in future)

TM import: ITD, zipped ITDs, MDB (SDLX TMs). This is a partial solution to the lack of TM Maintenance feature I’ve talked about in this article.

Continued lack of support for TMX is attributed to the fact that this open-source format has too many proprietary specifications.

Reporting formats added: CSV, Excel 2007, PDF, RTF, Word 2007.

Branding and Fonts are customisable (by Professional Services).

TMS 2010 is expected to have end-user customisable reports.

QA Models

This all-new feature in SP4 is crucial if your workflow includes Linguistic Review. All changes made by the Reviewers are now recorded, and the Reviewers can tag them using customisable Error Rating and Categories.

1. Error Ratings and Categories: support for LISA model, SAE J2450, TMS classic out-of-the-box.
2. User-specific models can be created. Number of points deducted can also be specified in the QA Model.
3. Records can be retained at segment (for feedback to translators) or project level
4. Scoring methods: absolute or percentage
5. To apply a QA Model: add it to a Configuration (i.e workflow), and it will be available to Reviewers working on jobs passed through this config.
6. Reviewer usage: click Star at segment level to open the QA model window and enter Category and Rating.Pass/Fail status does not prevent reviewer from submitting or rejecting a job.

memoQ 4: Interview with István Lengyel

I have been trying to diversify the topics we cover on LocLoc; and especially the tools we talk about. It started recently with a QA tool and now continues with a CAT tool. I already know from the survey I’ve had on this page, that a lot of you are familiar with Kilgray’s memoQ. This, is a preview of what to expect from the forthcoming memoQ4, from the mouth of Kilgray’s COO, István Lengyel.

[Nick Peris] Hi István, could you introduce Kilgray and your role within the company?

[István Lengyel] Hi Nick! Thanks for inviting me to do this interview. Kilgray Translation Technologies is an independent company dedicated to the development of clean and innovative tools for translation, but so far we are by far the best known for our memoQ translation environment. Though we are based in Hungary and all the founders are Hungarians, we became quite an international team in the last two years, opening up in Germany, Poland and now in the US. It’s really great to work in this team, as we have people coming from all sorts of companies such as Idiom, Passolo, SDL Trados, etc., and every addition to the team opens up new perspectives and shows new approaches – the company culture builds on respect and cooperation.

I am one of the architects of memoQ and also the chief operating officer at Kilgray, though in reality I’m mostly managing our sales and marketing team and our international expansion.

[Nick] Could you give a general overview of what memoQ is for readers who are not familiar with it?

[István] memoQ is an integrated translation environment that has a couple of focal points. First, it is easy to use, easy to learn. Second, we translate a lot in it and manage memoQ’s localization in memoQ itself, so we developed an eye for details – there are lots of smaller features that really make life easier. Third, from the very beginning we were concentrating on collaboration, and even the first version included an internet-enabled TM/TB server. Fourth, we don’t believe that we should lock in any of our customers – the entire system supports interoperability between tools to the maximum extent, meaning that you can process files prepared by virtually any major translation tool, and you can also prepare files for processing in other tools. There’s also a full set of documented APIs available for integration with other tools. Fifth, leverage, which means that we are trying to make the most of your resources. There were a couple of things where memoQ pioneered: we were the first to introduce real-time previews that change as you type, we were the first to introduce communication such as knowledge bases and instant messaging and offline synchronization into a translation memory server, we were the first to introduce the translation memory-based segmentation where pre-translation emulates the way your translators join and split segments, and we were the first to introduce the automated concordancing. But quite frankly, we are just as happy to take over things that work from other tools as we are to introduce new stuff.

[Nick] I know you are preparing to release a new version; could you give us a release date for memoQ 4?

[István] A few days ago we named January 31, 2010 for the release date, but I was reminded that it’s a weekend. So the first week of February. (Well, who cares about weekends? :))

[Nick] What are the main changes from memoQ 3.5 and main reasons to upgrade?

[István] There are so many changes that I can hardly list them! memoQ 4 is the first memoQ version that really focuses on project management. We like to build bottom-up and believe that an organization will only have a good experience deploying a tool if the translators like it, and we spent the last five years making the translators happy. So let’s start with the revolutionary feature: post-translation statistics. Imagine a situation where several people are working on the same set of similar documents, using a server-based translation memory. There can be a lot of fuzzy matches coming from the other translator’s translated entries, but so far there was no way in any tool to enumerate these matches, because the person who starts working later gets more matches than the person who is the first to start. memoQ 4.0’s post-translation statistics will solve this Gordian knot, and give you the actual fuzzy match analysis for every translator after the project. This way finally there is a business model for server-based translation.

Other than this, the biggest change is that we have upgraded the concept of translation memory servers to the concept of resource servers. So far you could share translation memories, term bases and documents between translators, and you could set up projects for them centrally. In the new version, you can share every other resource such as auto-translatables (for people used to Trados lingo: customizable placeables), non-translatables, segmentation rules, QA settings, keyboard shortcut settings, ignore lists for the spell checker and so on – 12 of them, all together. What’s more, sharing this happens in the background so you can start the publication of a big TM on the server and go on managing other projects in the meantime. These resources can all be exported into an XML-based format so clever project managers can prepare them also automatically.

memoQ 4 also brings finally the concept of multilingual projects. You can create handoff packages and receive delivery packages, or you can simply publish a project on the server. Those who receive the handoff package can in turn create new handoff packages (handy for a multi-tier enterprise-MLV-SLV-translator setup), and through delivery the files and reports are updated automatically. The handoff packages are just zipped containers of open-source format data – XLIFF for documents, TMX for TMs and CSV for terminology. You can process the packages in any tool, so the users are not locked in.

Compared to these improvements, the brand new text editor, the completely revamped user interface and the streamlined quality assurance seem small. Even the previous version of memoQ got quite a lot of credits for its good support of bidirectional and CCJK languages, memoQ 4 takes this further and also introduces support for Indic languages. We are introducing a very advanced multi-tier undo/redo logic, real-time spell checking and other minor improvements. The quality assurance checks have also been dramatically improved and also the interface for fixing warnings has been fine-tuned.

And I failed to mention so many things! memoQ 4 is the single biggest upgrade memoQ ever received.

[Nick] For non-memoQ users, could you give us the main reasons to switch to memoQ 4?

[István] Because other people do and they are happy about it! 🙂 Just like every company, we make mistakes at times but there has not been any single case that anybody asked for a refund. Seriously, I think the main reasons to switch to memoQ are collaboration, interoperability and support. memoQ is a truly collaborative application, it is one of the few tools that enable simultaneous translation and proofreading on the same document, complete configuration of projects for your translators, or using several translation memories or term bases that can be local, remote — they can even be on different servers — or offline synchronized. The server is fast even on a HSDPA connection and it’s also very affordable – no wonder we have over 150 servers out there.

The other important aspect is interoperability. Our main market is language service providers, and an LSP can never say that they use only a single tool, period, otherwise they lose business and what’s more, they can also lose translators. With memoQ you can process documents and packages created by other tools, and you can prepare packages in industry-standard formats for other tools too. Therefore you don’t find yourself in a situation that you bought the tool because you liked it and then you have to fight with everyone around you to make it accepted.

And the third most important aspect is support. I think Kilgray’s support is just great – fast, focused and friendly.

[Nick] What is the pricing structure for memoQ 4?
What are the different Editions of memoQ 4?

[István] memoQ 4 comes in three client editions: translator standard, translator pro and project manager.

memoQ translator standard is for those translators who never work in teams. It does not enable access to servers and does not enable export of files into XLIFF or bilingual DOC, only memoQ’s proprietary MBD format. It also lacks the ContexTM (101%) matching which takes the context also into account, and comes without support. But the price tag is attractive: 99 euros a year.

The memoQ translator pro is the edition for professional translators and very small translation companies who don’t want to invest into a server solution. It costs 620 euros.

The memoQ project management edition comes with multilingual project management and reporting functionality and we charge around a thousand euros for that.

When it comes to server technology, we sell our solution with mobile (ELM or floating) licenses, meaning that companies can give away and take back licenses to translators over the internet. The initial package contains five mobile licenses, and we sell additional bundles of five licenses at very competitive prices. When it comes to servers, we prefer not to sell without a trial period of 30 days – we want everybody to use the tool, not just buy it for the drawer.

[Nick] How did you take into consideration user feedback during the development of memoQ 4?

[István] Oh I could name the people who contributed with their user feedback here! I think it’s worth mentioning how we work. Basically there are four people who decide on what gets into the next release, and every release has a theme. These themes are contained in our 5-year roadmap and we regularly come together for things that we call “walk in the woods”‘ – creative sessions outside the office where we discuss the main ideas and concepts. We personally talk a lot with users and try to learn the rationale behind their feature requests. These talks shape the main themes/features a lot. On top of that, we have a system to archive all the threads on feature requests, and we go through these regularly. I could give you a rather precise list of features for the next three versions!

So basically the user feedback is taken into consideration on two levels: when we realize that a business problem is hard to solve with memoQ, we incorporate the solution into the high-level concepts. The other level is the feature level where for example users request amendments to file filters or suggest small usability improvements. If these are justified, these can go straight into the feature overview.

[Nick] How is Terminology Management undertaken in memoQ 4? What are the Termbase formats supported?

[István] Terminology management is one of the most controversial components in memoQ! So far we only support CSV and – surprise-surprise – TMX as import formats and can also export into Multiterm XML. Why TMX? Just think about software localization and then the help and you’ll understand. With memoQ we decided that this is a translation tool and not a terminology application, and therefore we gave a finite set of attributes but something that is pretty comprehensive: you can have synonyms, definitions, notes, grammatical information, contexts, project, domain, subject, client information, and a few other fields. You can also have images in the term base, and forbidden term variants can also be flagged. From the workflow point of view, memoQ has had a term base moderation feature since v2.0 in 2006, which means that terminologists may need to approve all terms suggested by translators before they become final. Terminology matching is really exciting: you can use wildcards to indicate the end of the invariable part of every word in a term, i.e. for a language like Spanish you can enter cinturón* de seguridad and that will also find cinturónes de seguridad. For translators of Slavic languages this is really crucial (fuzzy matching does not always work for terms). I can list quite a few pros for memoQ’s terminology management but I must say that it’s a very practical approach. However, we understand that corporate terminology management is not a subset of translation, and terminologists may need some more freedom.

Expect that freedom in a third-party tool based on the memoQ engine soon.

[Nick] Is there anything specific to memoQ in the way Translation Memories are created and maintained?

[István] Translation memories are by default context-enabled in memoQ, and memoQ supports two kinds of contexts: the segment before and after and context bound to structural information. This latter means that if you have for example the software strings in an XML or Excel file, with an attribute indicating where the text appears, you will get a 101% match if the attribute is the same to the attribute where you originally entered this translation – this way you can shuffle the translatable strings and still keep the context information. If you speak the Idiom lingo, this is very similar to ICE and SPICE matching.

As for maintenance, there are a couple of things that are quite unique. First, a 100% or 101% match for us is only a match that is identical both in content and formatting to the original. But we have a special bracket, 95-99% that contains segments where numbers, formatting, whitespaces, punctuation marks can be different. Any change in the text results in something lower than that. You can join and split segments wherever you want, and when you get an update to the document, the TM-driven segmentation will automatically join and split the segments according to your previous translation, as it looks into the translation memory for better matches through joining and splitting. During pre-translation, cases where you get multiple 100% matches (because you translated the segment differently in two contexts, and this third context is unknown so far) are flagged and they are very easy to locate. All these features fall under the umbrella term we use for design: “reproducibility”. I think it’s also worth mentioning that memoQ has a built-in TM editor and can work with as many TMs at a time as you wish. Oh, and yes, a minor nuance, just to make things elegant and please those who are really tech-savvy: our support for TMX also covers attributes, so if you import a TMX file coming from another tool that has attributes, even if the TMX attributes there cannot be displayed in memoQ, you can expect that the TMX export from memoQ will preserve and contain them – so memoQ does not swallow the information that it cannot process.

[Nick] Is there any new feature in memoQ 4 you are particularly fond or proud off? Maybe some anecdote about features which took you a lot of efforts to achieve and which you are now very happy to bring to memoQ 4 users?

[István] Well, I’m a person who prefers the big picture to the small details, and for me the biggest achievement – and a big praise goes to Gábor Ugray, our head of development who designed these features – is that the tool did not get more complicated for translators according to the feedback of those users whom we showed the system. We always pay a lot of attention to the user interface, but when we started conceptualizing memoQ 4 about two years ago, keeping its simplicity seemed like a daunting task. The visual marker of the entire resource management and multilingual project management feature is now just two drop-down lists: the server selector and the language selector. And I am of course proud of the fact that the resource concept makes the entire system future-proof – no matter what sort of a linguistic resource comes into existence in the next years, we’ve got a place for it, and savvy users are also welcome to write third-party resource managers.

[Nick] We are seeing a merging trend where tools are less specific to either software or documentation. This is partly due to the content types evolution, and partly to an effort by tool developers to become more all encompassing. How does memoQ fit into this? How is your support for software localisation? Also xml and xliff?

[István] I saw this very much in 2005 when we started off but I don’t see it that much anymore. About a year ago or so we implemented visual localization support for RESX files and quite a few users are using it, but we have no plans to implement visual localization for other formats such as RC or binary files. On the other hand there are quite a few considerations in memoQ that make it a very good tool for localizing Help content. I already mentioned the TMX import into the term base and the support for context based on another column in the Excel file or an attribute in the XML file, I’d like to mention the automated concordancing feature that was inspired by one of our translation jobs – in our earlier lives as translators – where TM management (another issue I could talk about for hours) was virtually non-existent. I don’t want to name the end-client and the LSP we got this from (they are both very reputable and well-known in localization), but basically to translate the help of version 8 of a well-known application we only got a TM that contained version 2 to 7 of the same application. No terminology, no localized software strings for version 8, nothing. We spent hours to find out what screen caption has been translated before and what expressions did we have to coin, because – as it is with software – quite a few of them were 8-10 words long, and of course developers make changes to these every now and then, changing one or two words maximum, adding a few words to the end, etc. The automated concordance automates this manual process: it automatically gives you the longest multiword expressions that appear at least a given number of times in the translation memory. It does not give you the translation in most cases, but if you select it, it opens the concordance window with the right expressions. And yes, the concordance can look for a series of words. So basically we don’t want to take away business from the excellent software localization tools, but we definitely want to be the best technology for translating help and manuals.

[Nick] Do memoQ and Kilgray offer workflow technology allowing supplier and clients in the localisation chain to work together online?

[István] Our workflow is a linguistic one, and not a highly structured one. We coined two terms. For us, horizontal workflow means when people work together on the same task. Vertical workflow is the traditional workflow, passing along the files between different people doing different jobs. memoQ is excellent in helping people work together on the same task and has a lot of workflow tools such as moderated term bases, simultaneous translation and proofreading, different forms of review, communication and knowledge bases, etc. From the point of view of traditional workflows, we only cover translation and review – items that happen within the tool. There’s no way to integrate things like source text review, DTP or settlements into memoQ. However, the extensive set of APIs enable integration with workflow tools, and at this point I have to mention that both Beetext Flow and Plunet Business Manager do a great job when it comes to deep integration. They can both take care of the entire process, and generate and maintain the projects automatically in memoQ. One of the things we are putting a lot of emphasis on nowadays is client review. I think memoQ is one of the best tools for this, but there is still a lot of room for improvement.

[Nick] Could you say a few words about the memoQ support network? How can new users avail of the experience of other users and if necessary receive support from Kilgray directly?

[István] Here are a couple of interesting resources: http://rc.kilgray.com – the Resource Center that contains training videos, guides, filter configurations for XML-based file formats, but also interesting articles on general topics such as TM management, technology purchase pitfalls, etc. for people and companies not using memoQ.

The memoQ Yahoo! Group (http://tech.groups.yahoo.com/group/MemoQ/) offers the expertise of other users but we also contribute often, and hey, you have the best experts of the competition also there and they often contribute too.

There is a memoQ wikibook too, and the forums on proz.com and other sites can also be interesting.

If direct support is required, it’s primarily through our support email address – please don’t publish the address directly on your website, we don’t want more spam there, but it’s at kilgray.com.

[Nick] Is it too early to ask you about roadmap? What are you plans for memoQ?

[István] It’s not too early at all, but I’m afraid I can’t tell much about the big improvements at this point. One thing is for sure – after 4.0, we will relax a bit and iron out any rough edges that may have remained in this brand new tool. One of the things that many users asked for and will be there in 4.1 (or whatever the final version number will be) is the bilingual DOC table format for review with comments. But one thing is for sure, you can expect another major version with a huge new resource in 2010.

[Nick] This has been a very informative interview. I thank you for your time and detailed answers and look forward to reviewing memoQ4 in the new year!

SDL Trados Studio 2009: Installation Guide

As previously announced, SDL released the new version of Trados late last week. To support the launch, the SDL Trados Support team have published some information about installation which I’m relaying here.

System Requirements

Pentium IV Dual Core
2GB RAM
Large screen resolution (1280*1024)
Windows XP SP2, Windows Vista SP1 (32-bit only for both)

(more details on the SDL Support site)

Preparation

1. Return your SDL Trados 2007 Suite Freelance (not Professional) activation code from the License Manager (view Activated License, select it, click Return License). This is a requisite before Trados Studio 2009 upgrade becomes available from your SDL account (due to the upgrade discount).
2. Uninstall beta/rc version (including manually for some files).
3. Uninstall any previous version of MultiTerm.
4. Ensure you have installed SDL Trados 2007 Suite (version 8.2.863 or 8.3.863). This is a pre-requisite to allow TM upgrade and support for ttx and itd in SDL Trados Studio 2009. A full and permanent version of Trados 2007 Suite is available from your SDL account if you have purchased or upgraded to SDL Trados Studio 2009. More info on this is available on the SDL Knowledge Base and ProZ.

Installation

1. Run the SDL Trados Studio 2009 installation (required .NET 3.5 platform included).
2. After completion, the SDL Product Activation wizard appears. It lets the user select between Buy, Purchased and Trial and then the license type (incl. License Server configuration).
3. Run the SDL MultiTerm 2009 Desktop (bundled with all versions of Studio 2009 – SDL MultiTerm 2009 Extract is sold separately).

Configuration

At first run you can select a Profile depending on your keyboard shortcut preference:

• Default: for new users
• SDL Trados: for users of previous Trados versions
• SDLX: for users of previous Trados versions

The rest of the configuration is automatic, and subsequent starts will be faster.

Notes

Winalign 2009 is not released yet. It will be provided for free to Studio 2009 license holders who in the meantime can use the version bundled with Trados 2007 Suite.

License server has not been updated, the existing version works with Trados Studio 2009.

AutoSuggest is included with every license of Studio 2009, but creating dictionaries requires a paying add-on (currently bundled for free for a limited time)

Alchemy Catalyst 8.0: Official Launch

On Friday, May 1st 2009, Alchemy Software Development officially launched a new iteration of their visual localisation tool and flag-ship product: Catalyst 8.0.

The event was held in Dublin (Ireland)’s Alexander Hotel, minutes away from Alchemy’s HQ. On offer were a feature highlights demo by Director of Engineering and Chief Architect Enda McDonnell, an informal meet-the-developers opportunity and client case studies by representatives of Citrix, Creative and Symantec.

This article reports and comments on some of what was said and shown.

A Total Visual Localization™ solution

Created mostly as a software localisation tool, Catalyst has now clearly outgrown this limiting description. The trademark visual editing capabilities now cover most aspects of localised content publishing:

• Help
• Web sites
• Software applications

Reaching out to translators

But Catalyst is sometimes still seen as an engineer’s tool. Alchemy are aware of this and have been listening to feedback from professional translators. The result is a translating environment which undeniably seems more linguist-friendly. There is a convergence with the interactive translation environment in Trados, which is only a part of a general strategy to increase translators productivity by lowering the time needed to get accustomed to various tools.

• Translator tool bar:
  • live validation: flagged with non-intrusive warning symbols
  • keywords: locking and validation for in-segment non translatables
  • internal tag management
  • multiple matches displayed
• Switch to the industry-standard terminology exchange format (TBX)
• Supplementary Glossary for translators to populate their own reference material
• Unlimited number of TM’s and web-based Machine Translation (MT) service ensure there is always a match

Changes to ezParse

In order to keep up with the long-served ambition of providing support for the latest file formats, changes have been made to Catalyst’s parsing tool.

• WPF (baml): full compatibility including visual editing of WPF forms and parsing out of.NET 3.0 objects
• Conditional XML: can now set the value of an element (or one of its attributes) to be localisable only if the value of another of its attributes indicates it should be treated as such (similar to functionality added to the settings file in Trados 2007).
  
• Multilingual XML: supported by reading the source segment in one element but storing the translation entered into another. While this is a very up-to-date feature, there seems to be some limitations in term of process. The translators will only deal with one language pair, so post-translation engineering will involve leveraging from multiple partially translated TTK’s back into the “Master” TTK before a fully multilingual file can be extracted. This should however be made easier by the updates made to Experts such as Leverage.

Updates to the Experts

• Programmable API’s (Com and Event) are provided to encourage client-developed automation. This was a strong theme across both the Alchemy presentation and most of the guest speakers’. It has been a feature of Catalyst for some time but is now emerging as the area where Catalyst gets ahead of the CAT pack.
• Multiple TTK’s, multiple languages and multiple TM’s to leverage from, all at once: this sounds like great news and is the feature I personally look forward to the most.
• Target folders can be set and original TTK’s preserved (necessary to achieve previous point).
• Leverage algorithm improved to search for 100% match in all TM’s provided before searching for fuzzy matches.

Cutting-edge Technology

• Improved navigation: thumbnails for Forms, Dialogs, WPF, HTML, graphics…are the latest addition to the visual features.
• Improved validation: live and programmable (API). Catalyst 8.0 comes with an updated list of validation tests and also offers the ability to create your own: custom .NET objects can be called by Catalyst during Validation but also file insertion, extraction etc.
• Underlying technology upgrades make Catalyst future-ready: compiler upgraded to Visual Studio 8 which is relevant both to Windows 7 compatibility and a future 64-bit Catalyst)

Screen caps courtesy of Alchemy

SDL Trados Studio 2009: Preview

Have you, like me, been slow to adopt Synergy? Do you maybe find it a little cumbersome or incomplete? Would you rather just open Workbench and TagEditor and get on with it? Or perhaps do you (or your clients) still find it easier to use Trados in conjunction with Microsoft Words?

Well, this may be about to change!

Earlier today, SDL conducted one of their very informative Webex meetings to announce a new version of Trados: SDL Trados Studio 2009. The release is due in June 2009, although “Trados 2009” is still in the last stages of development, so some of the features might yet change slightly.

The 1-hour short webinar comprised of 2 parts: a features highlight and a rapid but enlightening desktop-sharing software demo. Here is what I thought was worth bringing up to your attention: 
 

Feature highlights

Integration

This is actually quiet attractive and the reason why I brought up Synergy above. SDL seem to have come up with a truly integrated environment for editing, reviewing, terminology management, project management and all the aspects of Trados related work. No need to open a TM in Workbench, load a TermBase, open a TTX in TagEditor, a document in Word, or turn the coffee machine on.

Productivity

• New TM engine: the xml-based RevleX™. Among other things, it revives contextual match by liberating it from comparing old and new TTX files. Context Match works live, within any new document, and between files within a project.
• AutoPropagation™ immediately translates repeated strings within a document once you have translated the first occurrence.



• Searches can easily be run on both source and target segments.
• Multiple TMs lookup is available.
• AutoSuggest™: predictive text which leverages phrases rather than only segments from your TM as you type.



• Real-Time Preview: check final look as you translate, without navigating to a different tab. This seems very good news for those translators who find Trados tends to disconnect sentences from the whole document and lead translated documents to become a collection of sentences rather than a wholesome piece of work.
• QuickPlace™: improves text formatting, tags, placeables, variables management by providing it in-line.
• DTP application support has been updated and PDF can now be edited directly.

Open platform

• New XLIFF-based default format for bilingual files (.sdlxliff). Yes, this does mean the end of TTX files!
• Improved TMX and TBX support.
• Easy access to API for 3rd party applications.
• Customisable User Interface (UI).

 

Software Demo

As I mentioned before, SDL Trados Studio 2009 builds on Synergy. The interface has the now familiar Visual Studio .net feel which we’ve seen in Synergy as well as other CAT tools.

From the point of view of a Trados user, as in a Workbench + TagEditor user, the integrated aspect really becomes more prominent and inevitable, but in a good way!

Tab views

As expected in a Visual Studio.net application, a number of tabs are available at the bottom left of the UI. Some are familiar, some not:

• Start: provides the general overview.
• Projects: has new project status and Task History panels.
• Files: navigation pane has My Tasks and Sent Tasks folders to promote standardised filing.
• Reports: segment status.
• Editor: contains the entire interactive translation environment (more in the dedicated section below).
• TMs: preview, maintenance, update string, search from within the Trados Studio UI.

Editor

• A document can be opened from the main UI by simply clicking Open Document. But there is also a Windows Explorer context menu shortcut, which seems very efficient compared to opening Workbench, then TagEditor like you would most likely do with your current version of Trados.
• The Editor panel now has TM + Bilingual file+ TermBase + Previews all open at once.
• Source and target segments appear in a very clear and tag-free left-right panel view. This immediately seemed much more welcoming than TagEditor.
• Context Matches are flagged with a CM icon – not dependant on having a matching old ttx, also works live within new documents.
• Formatting can still be copied from source to target.
• Placeable and terms are offered in context (drop down like predictive text). No need to use arrow icons at the top of the UI (keyboard shortcuts still work).
• AutoPropagate seamlessly pre-translates further occurrences of strings you have just translated. They are marked as Unconfirmed 100% (orange instead of green).



• Full terminology functionality is also integrated, including adding to termbase.
• A Review mode allows to filter by match type (e.g. display only Unconfirmed 100% matches within a document for batch review and sign off).
• Editor can edit PDFs (but deliverable output isn’t PDF).

Project view (for PMs)

• Project templates can be saved with a high level of customisation.
• QA Checker is now in version 3.0.
• TM options can be edited from here.
• Dictionaries for AutoSuggest can be added.
• Tasks can be assigned to users during project creation. This information is then included when packages (i.e. translation kits) are created.
• Files can be merged, which creates a single .sdlxliff file out of potentially several file types.



• Batch processing: TM tasks are processed simultaneously (analysis, pre-translate etc.)
• Project package contents:
  
  • Can include Main (or Master) TM.
  • Can include an existing Project TM in a main package or create separate Project TMs if multiple packages (.sdlppx) are distributed.
  • Can link-up with Outlook to send automatically populated Handoffs emails.

  

• TMs view:
  • Can search through source and target.
  • Can upgrade existing TM.
• Requires all participant to be using Studio 2009

SDL Trados 2007: Translation Memory Strategies

What is the best way to organise and maintain Translation Memories?

I currently maintain TMs using 2 features of Trados (the Attributes and Master/Project TM dichotomy) and Alchemy’s Trados component.

Master TMs

• single and exhaustive repository for each field and language pair (e.g. EN-FR Medical).
• used to analyse all new projects and generate Project TMs.
• content of Project TMs are only added to it when full project cycle has ended (including review, QA etc.).
• because of their exhaustive nature, Master TMs tend to grow rapidely and would not be practical for inclusion into a translation kits.
• even when outsourcing all or most of the localisation process, these should always be held by the client as they are a valuable asset which they own, regardless of whether they are outsourcing TM Management.

Project TMs

• specific to a project or project stage (i.e. successive handoffs of a same project often have their own Project TMs).
• used to pre-translate the handoff (i.e. generate the TTX files to send to the vendor).
• passed on to translation vendors for analysis and use during interactive translation.
• used during post-translation engineering (bugs are fixed in Workbench + TagEditor + MultiTerm interactive translation environment by the localisation engineer).

Software TMs

• single and exhaustive repository for each field and language pair, generated bi-yearly from Catalyst TTKs.
• added to Master TM of their field and/or used as Concordance reference during translation of help, documentation, knowledge base articles etc.
• also used as leverage source for software through Catalyst.

Use of attributes

• every time a new project is analysed, custom attributes are added and set (e.g. Vendor=AAA, Project=XXX, Field=FFF).
• can be used to filter searches and analyses.
• also useful to track back on errors or arbitrate between duplications.