Wrapping up Locspire’s work with Naked Heart

Locspire, a localization team I was part of which was formed by 2nd year students of the Middlebury Institute’s Translation and Localization Management program had the great fortune of localizing content for Naked Heart, a non-profit with the very noble goal of building an inclusive society for people with special needs. Content was localized from Russian to English, Chinese, and Korean.

Naked Heart was founded by Natalia Vodianova in 2004. It was started with the goal of providing parks and playgrounds that are accessible to children with special needs. As of today, the program has built over 175 parks and playgrounds. Additionally, Natalia Vodianova started the program entitled Every Child Deserves a Family in 2011, which Locspire has been involved in localizing video content for.

While Naked Heart was already holding events internationally, they accepted Locspire’s help in localizing and translating content, developing terminology, as well as building a team of volunteers among other things. Naked Heart showed great interest in releasing more content in both English and Chinese, and Locspire helped them achieve this. In addition to Every Child Deserves a Family, Locspire worked to localize content for other special programs and events that Naked Heart either hosted or participated in.

Events and Programs Covered

• Running Hearts/Green Marathon
• Autism Awareness Day
• London’s Fabulous Fair
• Naked Heart Forum
• Every Child  Deserves a Family

Accomplishments:

Naked Heart already creates many promotional videos for the foundation in Russian, and as a member of Locspire, I had the great fortune of translating one of their most popular videos from Russian into English. This 20 minute long video explains the details of the new program Naked Heart has launched, “Every Child Deserves a Family”.

After this project was translated into English, our Korean linguists then translated it into Korean, and I along with another senior Russian linguist worked closely with the Korean translators to make sure the Korean subtitles were displaying at the appropriate times. While this was somewhat time-consuming, the results of the Korean subtitles were particularly stunning. A snippet of the video subtitled in Korean is posted below. You can also watch the full version subtitled in English here. Make sure you have closed captions turned on!

In addition to subtitling this lengthy video in English and Korean, Locspire worked on many other translation projects, the bulk of which were from Russian into English. Below is the list of projects that Locspire worked on along with their locales. In total Locspire completed 14 projects for Naked Heart (not including other services like terminology development)!

As briefly mentioned previously, Naked Heart had also shown a need for terminology development, and my team stepped in to fill this need by developing a multilingual (Russian, English, Chinese, and Korean) glossary for the foundation to pass along to both their translators and writers.

As the head Russian linguist I was mainly responsible for both translation and review; therefore, it only made sense for me to develop the Russian/English glossary. Below is a snapshot of it.

Locspire did not just help Naked Heart subtitle videos and release event materials. We helped grow a small but highly skilled team of volunteers to be used on special projects (such as video subtitling) in the future.  This includes translators, editors, reviewers, and specialists to help with desktop publishing issues. Furthermore, we developed an exit plan for them which involved some detailed suggestions about how to manage translation projects and independently interact with volunteers, as well as some other resources like glossaries and TMs (translation memories).

A Special Thanks to Naked Heart

Locspire would like to extend its gratitude to Naked Heart for giving us the opportunity to work so closely with an organization fighting for such a pure cause– an inclusive society for children with special needs. Thank you, Naked Heart, for allowing us to contribute something to this noble effort.

My vendor team, LingoTrans, was tasked with coming up with a proposal for a potential client that needed some translation work done. We were in charge of localization engineering and the translation of software/UI/metadata – .resx files, as well as: Marketing and Transactional Emails  (Word docs); Legal (Privacy Policy, Terms & Conditions, Cookie & Tracking Policy) (Word docs);Customer Service Content (pre-written answers for Customer Service, audio recording of after-hours phone message etc.) (Excel & mp4 files; SEM: Keywords, Google Ads, Banner Ads for GDN – Excel; SEO Content – Excel; Social media posts – Excel. We were also responsible for the initial roll-out and testing.

Overall, our RFP went well. We were able to win the favor of the potential client over another competing group. We were able to provide adequate answers to all questions asked to us by the client, and we also presented some solid calculations to the client on the cost of the entire project, and they did not seem put off by the numbers.

I think overall, our presentation of our timeline did not go as well as planned. While we initially planned to launch in May as directed in the assignment; however, during our pitch the client admitted that we potentially had as long as 6 months, which threw us a bit off guard and we had to regroup and essentially tell them that it could be done sooner than that,

There could, and probably should have been more of a focus on the numbers. At the very end of the presentation we gave a brief rundown of the cost of the project, and while we feel our estimate is close to the actual cost, we could possibly have given a more precise figure had we reserved more time for presenting them.

The lessons learned for this presentation are as follows: Nail down your timeline, and also consider what would happen if this timeline was changed significantly. Also, go further than just giving an estimate, if you can, give yourself more time in the pitch to offer some more detailed figures

Linguistic Review

Our client’s in-country employees (marketing team members, customer support, etc) are giving us feedback on translations. However, the reviewers frequently introduce spelling errors, terminological inconsistencies, and regionalisms, all of which conflict with the TMs, glossaries, and style guides. The marketing manager of the internal review team thinks that we are just being stubborn and inflexible, but that is not the case! We are only trying to maintain quality, while still valuing the opinions of our client’s team that is conducting the internal review. 

My team is tasked with introducing an all in one solution to the company that they will readily accept. There are a few challenges in this: the solution has to be easy, comprehensive, and more importantly we have to convince the company that this change is worth their time and won’t hurt their pocketbooks.Our main evangelization strategies include stressing the free and easy nature of our solution. We also stressed that training would be provided and gave an in depth review of the proposed technology while highlighting each of its functions that would solve the present issues.

What went well in the presentation: overall we sold our solution pretty well, and our client did not protest too much at the idea of having to use a new software, SmartCat. I feel like we were able to provide specific reasons for needing to use SmartCat, and additionally we were able to show that the solution we are proposing, in fact, has built in functions (glossaries, translation memories, due date function that automatically adapts to the time zone, chat system/ability to comment) that handle all of the problems that have come up when our team is incorporating internal review into the final translation. We were able to show the client this easily by providing several screenshots of the interface and giving a brief overview of the main functions before we go into more detail about how we plan to use this to solve the present issues. 

What went wrong:  Personally in my portion of the presentation, I feel that the main thing that went wrong was leaving out some points that I initially wanted to mention. Despite the information, which was just a simple explanation of the progress bar that SmartCat’s interface has as well as some other similar details, I started to rush myself and ended up omitting these important details. Also, I feel like my team forgot to leave out something very important, which is an alternative solution. While we discussed this solution, which is described in greater detail below. While we discussed alternatives in our group meeting, we never thought to include them briefly in the presentation, which as an after thought, would have been a good idea. 

What could have been done differently: I feel that enough time was not spent showing the client the (worse) alternatives which would have highlighted the brilliance of our solution. Some of the alternatives would’ve been using a CAT tool such as MemoQ, exporting the documents for review, and then retrofitting those changes again using MemoQ. Perhaps explaining this process and the cost associated with MemoQ licensing would show our client that we are serious about helping them in the most cost-efficient way possible.

Lessons learned: Any person who has been through formal education has heard the advice of not rushing through presentations; however, I did not heed that advice in the moment of my presentation, and this experience was a gentle reminder to take the time to deliver the information points you practiced earlier. Instead of rushing because you are worried about time, trust in the planning and timing that you practiced in advance. Give yourself time.

Another lesson learned would be to always present the alternative to your client, even if it is a bad alternative. If anything this will highlight the strengths of your solution and also show that you are flexible and open to other ideas if for some reason the client does not care for your preferred solution. I also think it is important to not present the alternative in a totally negative light (it is still your idea after all) and if the client ends up liking the alternative more than the original proposal, that could make both you and your client feel foolish.

Conclusion: While there were no major mistakes made during the presentation, and my team was easily able to win over the sharks, there is always room for improvement, especially in the areas of timing and presenting other solutions to the problem.

My Team:

During my localization practicum at the Middlebury Institute of International Studies I have had the opportunity to showcase both my knowledge of translation and localization by teaming up with four other students to create the team Locspire, which is currently working with the non-profit foundation Naked Heart. The senior members of the team are Russian, Chinese, and Korean translators who also have the knowledge and experience to double as localization specialists. Other members consist of junior translators who are involved with translation, editing, and review. I am a senior Russian translator and localization specialist.

The Client:

Naked Heart is a Russian based organization which is aimed at helping children with special needs. It was started by Natalia Vodianova in 2004. It was started with the goal of providing parks and playgrounds that are accessible to children with special needs. As of today the program has built over 175 parks and playgrounds. Additionally, Natlia Vodianova started the program entitled Every Child Deserves a Family in 2011, which Locspire has been involved in localizing video content for.

Project summary:

While Naked Heart is already holding events internationally, they have accepted Locspire’s help in localizing and translating content, developing terminology, as well as building a team of volunteers among other things. Naked Heart has shown great interest in releasing more content in both English and Chinese, and Locspire is helping them achieve this.

Accomplishments to Date:

Naked Heart already creates many promotional videos for the foundation in Russian, and as a member of Locspire, I personally had the great fortune of translating one of their most popular videos from Russian into English. This video explains the details of the new program Naked Heart has launched, “Every Child Deserves a Family”. The Russian version is embedded below, and the English version will be posted as soon as it is finalized.

Additionally, since Locspire has started working with Naked Heart, they have been able to publish a press release for a London based event in simplified Chinese and also follow up with a post release. The first page of the Chinese press release is shown below, and Locspire’s team of junior Chinese translators are credited with this translation.

Working towards the Future:

Locspire has not just been helping Naked Heart subtitle videos and release event materials. We are continually helping them grow a team of volunteers to be used now and in the future.  This includes translators, editors, reviewers, and specialists to help with desktop publishing issues. Furthermore, we are developing a consultation plan for them which will involve some detailed suggestions about how to manage translation projects and interact with volunteers.

Naked Heart has also shown a need for terminology development, and my team is stepping in to fill this need by developing a multilingual (Russian, English, Chinese, and Korean) glossary for the foundation to pass along to both their translators and writers.

Challenges:

Currently, we have been experiencing challenges related to communication both within the team and between our client. Right now all email communications made with the client are in Russian, so the rest of our team is relying on the two senior Russian translators to relay important information. In general, it is hard to keep everyone “in the know” via email, and threads quickly get out of control so I feel that internal communications could be more streamlined by using a service like Slack instead of email.

In the vein of communication issues, we have had some struggles with getting our client to grant us timely access to certain client-side systems (for example, access their subtitle editor) to finalize the translation of the video. This, unfortunately, forced us to use a more undesirable workflow when producing the English translation of the video, but it has also highlighted the need to educate our client about proper workflows so that this issue can be avoided when we receive another subtitling project.

Conclusion:

Despite the previously mentioned communication issues with the client, we are encouraged by recent messages from Naked Heart promising timely access to their subtitling editor and also more projects for translation! They have communicated that they have been very busy hosting event after event, and Locspire is excited and prepared to help them with localizing the materials that are undoubtedly coming about due to these activities. Upcoming Projects include translation of three news articles from Russian to English for Naked Heart’s website, as well as the translation of slides for one of the Foundation’s presentations.

I recently attended a workshop on crowdsourcing solutions for translation at the Middlebury Institute of Intl. Studies at Monterey. During the course of this workshop, we discussed two main issues with crowdsourcing that come up in translation: quantity and quality. This blog will deal with the latter, and is a follow-up to my blog on quantity.

Many organizations may be wary of crowdsourcing translations because they do not want sensitive or public facing documents to be poorly translated. While these fears are not completely unfounded, I believe that a high-quality translation can indeed come from crowdsourcing or community translation if some measures are put in place to ensure quality. Here is my list of the15 best general practices for controlling quality when crowdsourcing translation.

1. Develop a heavy screening/vetting process.

• Demand at least 2-5 years of experience.

• Ask for proof of any accreditations.

2. Consider a fast track to streamline this process. For example, persons holding an MA in translation or ATA certifications would receive immediate approval.
3. Do not allow volunteers to translate into their ‘B’ language.
4. Advertise yourself as a place where the translator’s skills matter.
5. Create a list of banned words for each language, and use content filtering.
6. Create a simple style guide.
7. Ensure both editing and review is done.
8. If human review is impossible, use an automated review tool to check for simple typos.
9. Ban use of MT.
10. Limit volunteers to only one language pair i.e. no ‘C’ languages.
11. Provide terminology.
12. Train and educate volunteers on how to use the organization’s tools.
13. Invest in volunteers; ensure they are growing.
14. Publicly recognize your translators so they are incentivized to produce high-quality work.
15. TRUST your community of translators.

Please also see the infographic I created on the 5 best practices for both quantity and quality, which combines the most important aspects from each blog on quantity and quality:

5 Best Practices for Both Quality and Quantity in Translation Crowdsourcing

I recently attended a workshop on crowdsourcing solutions for translation. During the course of this workshop, we discussed two main issues with crowdsourcing that come up in translation: quantity and quality. This blog will deal with quantity, and a follow-up blog on quality will be posted. 

Many organizations have such large volumes of content that paying for it is simply not feasible. Lots of organizations are paying for only a small portion of their content to be translated, and some organizations have started making progress towards fitting the translation of ALL of their content into the budget, and not just some, by using crowdsourcing. This is just one scenario where “quantity” may be one of the main goals of the crowdsourcing since other translations that need to be high quality are done by paid professionals. I have come up with a list of the 15 best practices that any organization could use for achieving maximum quantity when crowdsourcing translation:

1. Develop a leaderboard.
2. Use gamification techniques.
3. Use MT and TM. (Prepopulate MT or show as a suggestion)
4. Make the organization easy to join.
5. Open your doors to more non- professionals needing experience.
6. Allow translation into B language.
7. Allow multiple language pairs.
8. Provide field-specific terminology (which also helps encourage novices to try to volunteer).
9. Use a content filter for banned words.
10. Skip review for translation done into A language.
11. Skip review for “approved” translators.
12. Send notification reminder to translators.
13. Hold a remote Translatathon after recruiting, before translators lose interest.
14. Use gift card incentives.
15. Prioritize translations using a CVI (content value index).

Please also see the infographic I created on the 5 best practices for both quantity and quality, which combines the most important aspects from each blog on quantity and quality:

5 Best Practices for Both Quality and Quantity in Translation Crowdsourcing

Localization of Hangman in Visual Basic

In my Software and Games Localization class at the Middlebury Institute of International Studies led by instructor Max Troyer, I have gained a basic understanding of software and games localization. In my basic understanding, the localization of software or games can be broken down into three steps: engineering/internationalization; TEPR; QA/Testing.  Finally, this process results in all of your translations being integrated into the actual software or game. We are essentially extracting strings for translation, translating them, and then feeding them back in to our software program or game, using different tools and developing different processes during engineering as necessary, as games are unique. Of course, we can’t forget the important role that testing (both before and after localization) plays in this entire process.

My group chose to localize Hangman in Visual Basic for two main reasons

1. We really liked the game.
2. We knew it would be a challenge

We knew it would be a challenge because we had never worked in Visual Basic before. We had localized our own custom form in C# before, but never in Visual Basic. So we knew that there would be a few challenges ahead just because this is a new language for us to work in.  We also decided to add something extra to increase the difficulty of the project a bit. We added a language picker. For the purposes of this project we used machine translation.

Localization into Chinese itself posed an interesting problem because the nature of Chinese characters does not work in a traditional Hangman game. So in this case we decided to use pinyin, a Romanized version of Chinese to get around this issue.

In my group of three, I was responsible for designing the workflow of localizing Hangman in Visual Basic into Chinese, Spanish, and Russian.

The workflow is as follows:

Workflow:

• Play original game, identify strings for translation (labels/buttons on the interface, the Information pop-up, other system messages when entering a word, winning a game, losing a game)
• Implement resource manager/WinFormStrings method: create a .RESX resource file for English, wrap strings (in class we worked with C#, this is Visual Basic but the site we referenced in class https://msdn.microsoft.com/en-us/library/y99d1cd3%28v=vs.100%29.aspx?f=255&MSPPError=-2147217396 shows how to do it for both languages)
• Translate strings in CAT tool. MemoQ doesn’t support .resx files, but Trados Studio 2015 and Memsource do
• Make sure the new, translated files are properly named with the correct language culture names, put them in the project https://msdn.microsoft.com/en-us/library/ee825488.aspx?f=255&MSPPError=-2147217396
• Translate the word list
• Implement the language-specific keyboards
• Implement language picker (we found an example one online and modified it a little)

Screenshot of game before a locale is selected or a new game is started by the player

Screenshot of Pinyin/Chinese game. Notice how there are no center tiles for the hidden word appearing. This is a glitch we encountered.

Unfortunately, we did have one pretty major bug with this game that could never be resolved. Simply put, our bug was that the center tiles would not appear after a new game with a different locale was launched.  (see second screenshot). My team and I reached out the developer of the game to see if he could shed any light on why we were experiencing this issue, but even he was stumped. My team and I plan to continue working on this project in the coming weeks so we can resolve this bug, and this blog will be updated with the results of our work.

The link to the project template for the XML filter assignment can be found here:

https://www.dropbox.com/s/f3tdrnci0m7vl3t/WP%20Template.sdltpl?dl=0

Presentation on SMT