IMUG Past Events Archive: 2016
2016 Events:
- Dec: Annual IMUG International Potluck & Holiday Bash
- Nov:The China i18n Challenge: Western #Fails in Type, Typewriters, & Computers
- Oct: International Product Excellence at Google: from Inception to Post Launch
- Sep: How PayPal launched an app in 140+ countries simultaneously
- Aug: GALA-IMUG Networking Night
- Jul: Facebook Community Localization
- Jun: Localization for the eBay Global Marketplace
- May: Spoken Language Translation: Past, Present, and Future
- Apr: Serge: A Free, Open Source Platform for Continuous Localization
- Mar: Demystifying Unicode: the Fundamentals of Encoding a Script
- Feb: Translation Markup Language & Universal Translation Memory
- Jan: Simultaneous Interpretation for Virtual Meetings
Want to Speak at an IMUG Event? Want to Suggest a Speaker?
To propose a presentation for an upcoming IMUG meeting, to suggest a speaker for us to invite, or to get information on how to prepare, please see Information for Presenters.
2016 Events Archive
December 8, 2016, 6-9 PM
(This one is on the second Thursday, and starts at 6 PM.)
Annual IMUG International Potluck & Holiday Bash
Hosted by Adobe Systems in San Jose, CA
Photos: https://www.meetup.com/IMUG-Silicon-Valley/photos/27476301/
The IMUG holiday bash always features unusual food and good conversation, including discussions of language technology, business, travel, life across cultures, and more.
Dr. Ken Lunde and the Adobe globalization team has once again graciously offered to host our event, this time at Layers Café in the same East Tower as most of our IMUG talks.
Adobe will supply the drink including soda, wine and spirits.* IMUG will supply free pizza. You needn't bring a thing, but we always look forward to the fantastic variety of ethnic dishes the less culinarily-challenged among us can muster up for this annual food-fest. That always takes things to a higher level.
You needn't prepare or buy more than 3 to 4 portions worth, and in fact you needn't bring anything at all. Come as you are, with or without a potluck contribution, but do come prepared to eat!
November 17, 2016, 7-9 PM
(Networking 6:30-7:00 PM, and after the talk as time allows)
The China i18n Challenge: Western #Fails in Type, Typewriters, & Computers
Hosted by Adobe Systems in San Jose, CA
Photos: https://www.meetup.com/IMUG-Silicon-Valley/photos/27432087/
Videos: Dr. Tom Mullaney at Google on Chinese typewriters and the birth of predictive text and at the Computer History Museum on Chinese input methods, privacy, and surveillance in the cloud. We regret not recording his IMUG talk, but perhaps he'll be back to tell us about his next book soon!
Ever since the mass manufacture of keyboard typewriters began in the United States, companies like Remington and Underwood imagined a day when this new device would conquer the Chinese language in the same way it had practically every living script across the world. It never did.
Ever since the invention and popularization of hot metal printing in the United States and Europe, companies like Linotype and Monotype dreamt of a day when their technologies would conquer the Chinese language, just as they had Arabic, Armenian, Burmese, Devanagari, Hebrew, Korean, and over one hundred other scripts. They never did.
Ever since the advent of personal computing, companies like IBM and others imagined a day when Chinese keyboards would behave "just like ours," even including the familiar QWERTY keyboard. They never did (even though QWERTY is now ubiquitous).
Drawing upon over a decade of research on modern Chinese information technology, Tom Mullaney of Stanford University will reflect upon the vital importance of history and culture in the process of technological internationalization, particularly in China.
Dr. Mullaney is the author of two forthcoming books from MIT Press, The Chinese Typewriter: A Global History of the Information Age, Part I, to be published next year, followed by The Chinese Computer: A Global History of the Information Age, Part II.
October 20, 2016, 7-9 PM
(Networking 6:30-7:00 PM, and after the talk as time allows)
International Product Excellence at Google: from Inception to Post Launch
Hosted by Google in Mountain View, CA
Photos: http://www.meetup.com/IMUG-Silicon-Valley/photos/27362232/
We hope to post a video soon too!
Join us to find out how the Google Localization team is tackling the challenge of ensuring products feel native in 75+ languages. We will discuss key challenges and solutions that result in excellent products that delight global users, from product design and linguistic quality frameworks to post launch feedback.
Mackenzie Nicholson is a Solutions Program Manager for Localization at Google. Prior to Google, she worked in international product for an e-commerce company. She is passionate about connecting people through shared products and experiences that transcend borders. She is also an efficiency junkie who loves simplicity, optimization and scaling.
Stephanie Dupuy is a business unit lead on the localization team. In this role, she oversees the localization for Android, Chrome, Play, YouTube & Hardware products. Prior to Google, Stephanie was a management consultant specialized in high-tech & banking. She holds Bachelors and Masters degrees in international business and international management from ESSEC (Paris), UCD (Dublin) & Bocconi (Milan). She is passionate about creating products that delight users globally!
Additional panelists from the internationalization and localization teams will join in for further discussion plus audience Q&A.
September 15, 2016, 7-9 PM
(Networking 6:30-7:00 PM, and after the talk as time allows)
How PayPal launched an app in 140+ countries simultaneously
Hosted by Adobe in San Jose, CA
Video: https://my.adobeconnect.com/p56hcpamg2h/
Photos: http://www.meetup.com/IMUG-Silicon-Valley/photos/27276135/
These days most commerce happens online, and mobile has become a key player. This talk will explain how PayPal was able to launch the new PayPal Mobile app for iOS and Android simultaneously in more than 140 markets.
This was done using new platforms and process to make it easier to localize and customize. From design to development, the PayPal globalization team has been involved since day 1, providing internationalization support and QA, reviewing content, and delivering translations.
Lucas Welti is a Globalization Architect at PayPal, responsible for working with the next generation of mobile apps. He also works to evangelize i18n and L10n best practices across all PayPal products. Lucas joined PayPal in 2015.
August 18, 2016, 6:30-9:30 PM
GALA-IMUG Networking Night
At Steins Beer Garden in Mountain View, CA
Photos: http://www.meetup.com/IMUG-Silicon-Valley/photos/27205030/
IMUG is proud to once again partner with GALA on a local networking evening! Come for a tempting buffet, cash bar, and fine conversation with your GILT industry friends and colleagues on the private patio at Steins Beer Garden in Mountain View.
Free public parking is available at surface lots or parking garages less than one block in any direction, and we'll be just a few blocks from the Mountain View CalTrain station too. Look for the intersection of Villa and Bryant on this map: http://mountainviewdowntown.com/parking-map/
GALA local networking events are intended as relaxed, non-commercial gatherings. This is one of the many ways GALA carries out its mission to bring the industry together to share information, foster innovation among GALA members and the industry as a whole, and to offer clients collaborative value.
The Globalization and Localization Association (GALA) is the world's largest trade association for the language industry with over 400 member companies in more than 50 countries. As a non-profit organization, we provide resources, education, advocacy, and research for thousands of global companies. GALA's mission is to support our members and the language industry by creating communities, championing standards, sharing knowledge, and advancing technology. For more information: www.gala-global.org
*Please drink responsibly, and choose a designated driver or take alternate transportation if necessary.
July 21, 2016, 7-9 PM
(Networking 6:30-7:00 PM, and after the talk as time allows)
Facebook Community Localization
Hosted by Facebook
Photos: http://www.meetup.com/IMUG-Silicon-Valley/photos/27168198/
Over 1.6 billion people use Facebook each month to connect. Come learn how the Internationalization team at Facebook is scaling to meet the language needs of its many users. This talk will cover Facebook's Community Localization project and how the Facebook Internationalization team is making the world more open and connected in languages people understand.
Hajnalka Sarvari is currently leading the Internationalization Program Management team at Facebook. She joined the company in 2013 working as the Program Manager for Community Localization before stepping into her current role. Prior to Facebook she worked in multiple roles in the localization industry on both the supplier and customer side. She is deeply passionate about enabling people to connect by helping to eliminate the language barrier, and providing great language experiences. She is originally from Hungary, and currently lives in the Seattle area.
Aaron Babst is the Program Manager for Community Localization at Facebook. He previously spent several years at Google working on the Google Translate Community project and crowdsourcing on Google Maps. Aaron believes that when people are more open and connected it helps us embrace the differences of others and makes the world a better place. Aaron is originally from the Portland-area but now calls San Francisco home.
June 16, 2016, 7-9 PM
(Networking 6:30-7:00 PM, and after the talk as time allows)
Localization for the eBay Global Marketplace
Hosted by eBay
Photos: http://www.meetup.com/IMUG-Silicon-Valley/photos/27052140/
Slides: Localization-for-the-eBay-Global Marketplace-slides-only.pdf
Video: Coming soon!
Sabine Rioufol, Herbert Hoisl, and Marcel Bregman, representing the core functional groups with the eBay localization team (translators, project managers, and engineers), will give an overview of the processes and best practices eBay has put in place over the years, to address some of the challenges commonly faced in enterprise localization. In addition, each speaker will highlight a particular solution related to their specific area, focusing on how it helps optimize for velocity, quality and/or cost for a given support use case, for instance agile web development.
About the speakers:
Sabine Rioufol is the Senior Language Specialist Manager at eBay. She has been involved in the localization industry for 20+ years. She started her career as an in-house linguist at a small LSP, and has worked as a translator, interpreter and manager in various companies, including eTranslate. She joined eBay in 2003. where she now leads a team of 13 linguists – 9 involved in the traditional localization track and 4 involved in the machine translation. Sabine is passionate about establishing processes and recruiting, and people management to ensure the best quality. Sabine holds a Master’s Degree in Translation and Interpretation from Monterey Institute.
Marcel Bregman is the Senior Localization Engineering Manager at eBay. He has been active in localization for almost 20 years, and has been with the eBay localization team since 2001. During this time Marcel has been involved in localization process engineering, setting up l10n support for subsequent generations of the eBay platform stack, implementation of a Translation Management System, migration to a simship release model, internationalization education, among others. Marcel holds a Master’s degree in Chinese Studies from Leiden University, in the Netherlands.
Herbert Hoisl is the Senior Localization Program Manager at eBay. He has been active in localization for more than 20 years, and has been with the eBay localization team since 2003. He heads up a team of localization project managers, responsible for project planning and execution, client communications, vendor management, as well as metrics tracking. Besides managing his team, Herbert focuses on process improvements, planning of strategic initiatives, resource planning, client consultation, and budget management. Prior to eBay, he worked at Micrografx, Corel Corporation, and HighTech Passport Ltd.
May 19, 2016, 7-9 PM
(Networking 6:30-7:00 PM, and after the talk as time allows)
Spoken Language Translation: Past, Present, and Future
Hosted by Adobe
Photos: http://www.meetup.com/IMUG-Silicon-Valley/photos/27051122/
Video: https://my.adobeconnect.com/p79mnizx8uf/ (Sorry about the echo; it gets tolerable and this is a great talk!)
This talk will offer some history of the field of automatic speech-to-speech translation (combining speech recognition, machine translation, and text-to-speech); give a quick tour of the state of the art; and speculate about future developments, with special interest in deep semantics.
Dr. Mark Seligman is Founder and President of Spoken Translation, Inc. From 1983 until 1991, he was lead software trainer with IntelliCorp, Inc., a developer of artificial intelligence programming tools. In parallel, he pursued doctoral studies in linguistics at UC Berkeley, with a 1991 dissertation on the automatic generation of multi-paragraph discourses.
After graduation, he spent three years at ATR (Advanced Telecommunications Research) Institute International, Japan’s premier communications consortium, studying numerous aspects of speech-to-speech translation. Back in the U.S. from 1995, he continued his academic career as a Research Associate at GETA (the Groupe d’Étude pour la Traduction Automatique) at the Université Joseph Fourier in Grenoble, France, and as invited advisor at DFKI (Deutsches Forschungszentrum für Künstliche Intelligenz) in Saarbrücken, Germany.
In 1997 and 1998, in cooperation with CompuServe, Inc., he organized the first automatic speech translation system anywhere demonstrating broad coverage with acceptable quality. From 1998, he was Publications Manager with Inxight Software, Inc., commercializing linguistic and visualization programs developed at PARC. In 2002, he left to establish Spoken Translation, Inc.
April 21, 2016, 7-9 PM
(Networking 6:30-7:00 PM, and after the talk as time allows)
Serge: A Free, Open Source Platform for Continuous Localization
Hosted by LinkedIn
Slides: IMUG_Serge.pdf
Documentation: https://serge.io/
Download: github.com/evernote/serge
Photos: http://www.meetup.com/IMUG-Silicon-Valley/photos/26913023/
Video: https://youtu.be/bC3wECRgLog
Evernote is being translated into more than 26 languages, and although we have a very agile development process, we still manage to provide translations on time and ship Evernote releases simultaneously in multiple languages.
How do we do that? The answer is continuous localization. It is a fully automated way of dealing with localization, and it revolutionizes the localization process the same way automatic build integration and automated QA revolutionizes software development.
We recently open-sourced our continuous localization platform (it is called Serge, see https://serge.io/). In this talk, we will describe what Serge is about, how it is used at our company, and give you ideas on how you can use it for your own purposes.
About the speaker:
Igor Afanasyev is a Director of Localization at Evernote. During his 9 years with Evernote, he built localization infrastructure from ground up to meet the needs of the the rapidly expanding startup environment. In his current role, he supervises internationalization and localization at Evernote, continues to support and improve localization tools; Igor and his engineering team actively contributes to open-source localization community. Igor holds a bachelor degree in Engineering from the Moscow State University of Printing (Moscow, Russia), and has 20+ years of experience in software engineering and user interface design.
March 17, 2016, 7-9 PM
(Networking 6:30-7:00 PM, and after the talk as time allows)
Demystifying Unicode: the Fundamentals of Encoding a Script
Hosted by Adobe
Preliminary Lampung encoding proposal: http://www.unicode.org/L2/L2016/16073-lampung.pdf
Example of another script encoding proposal: http://www.unicode.org/L2/L2016/16068-vatteluttu.pdf
Photos: http://www.meetup.com/IMUG-Silicon-Valley/photos/26828910/
The ability to internationalize, localize, and translate content requires support for languages and scripts at the most fundamental software layer. Today, this layer is the Unicode standard. Encompassing more than one hundred scripts and substantial locale data, Unicode enables user communities around the world to create and consume content in numerous languages. But, how to support orthographic requirements when the basic technology that enables such support does not itself exist? The first step, of course, is to implement support for the script in Unicode.
This talk demystifies the process of transforming an analog script into a digitized encoding. The Unicode code chart that presents the encoded repertoire of a script is the result of a standardization process for which best practices or packaged solutions do not exist, but we will learn how this can be done by evaluating two cases. The first is a historical southern Indic script known as ‘Pallava’, which travelled to south-east Asia, Indonesia, and Philippines, and is the ancestor of the major historical and modern scripts of these regions. The second is Lampung, a modern minority script used in Sumatra, Indonesia, which is a descendant of Pallava. These two scripts are not yet encoded in Unicode, and they present several issues for standardization.
The talk draws upon analyses of Pallava and Lampung sources in order to explain decisions that inform the definition of encoding models, character repertoires, and representative glyphs for scripts. It also presents the technical issues regarding the encoding of these to illuminate various Unicode principles, such as the character-glyph model and unification.
Anshuman Pandey is a technology analyst, Asia specialist, linguist, and historian. Most recently, he was a post-doctoral researcher in Linguistics at UC Berkeley, where he was funded by a Google Research Award. He earned a Ph.D. in History from the University of Michigan. Anshuman has encoded more than twenty scripts in Unicode on behalf of the Script Encoding Initiative and is pursuing twenty more. He was awarded the 'Bulldog Award' by the Unicode Consortium in 2011 for his contributions to Unicode.
February 18, 2016, 7-9 PM
(Networking 6:30-7:00 PM, and after the talk as time allows)
Translation Markup Language & Universal Translation Memory
Hosted by Google
Video: https://youtu.be/5rR_Vi5TH2U
Photos: http://www.meetup.com/IMUG-Silicon-Valley/photos/26753562/
Today’s web and mobile app localization industry relies on numerous standards, libraries and file formats to facilitate the exchange between developers and translators. While some formats are somewhat sophisticated, others lack even the most fundamental features, like pluralization and contextualization. And most can’t offer support for more advanced localization features, like language cases.
The most common formats used today include Gettext PO, Android XML, YAML, .Net RESX, iOS Strings. Those standards are typically packaged with popular frameworks and in some cases leave the developer with no choice but to use them. A typical developer today works with many frameworks - for instance a Rails backend app (YAML) with Ember front end (i18n JS) and iOS mobile app (Strings). Since all standards have distinct syntax - in many cases translations cannot be shared across applications.
Translation Markup Language (TML) aims to solve both these problems by introducing a powerful extensible cross-platform syntax that offers support for pluralization, language contextualization, language cases, reusable decorators and much more. TML libraries are available for all major web and mobile platforms. TML allows translators to do in-context translations - where they can translate right from within the apps. TML libraries also eliminate the need for developers to ever deal with the resource files, as all extractions and translation substitution is done realtime and the resource files are only used as a transport between the apps and the Translation Exchange platform.
Translation Exchange stores all translations in Universal Translation Memory (UTM), a graph database which stores all translations with their context, tone, rank and other attributes for accurate matching. This allows translations to be shared across all apps in the Translation Exchange Network. The translation memories of each app are extracted from the UTM graph and are managed by their individual localization teams.
During this presentation we will look at some of the features of TML and how it can be used to quickly translate an app into a number of language using in-context translation tools. We will also look at how the data is stored and shared across applications using UTM.
Michael Berkovich is a co-founder at Translation Exchange. Before starting "TrEx", Michael was a lead engineer at Geni, where he built an open-sourced localization platform called tr8n. The platform allowed Geni users to submit, review and validate translations. Geni.com was translated to more than 20 languages in less than 3 months. The site is now available in more than 40 languages. Michael helped integrate tr8n at Yammer, Kongregate and other companies around the world. Before Geni and Yammer, Michael was a lead engineer at SOA Software (Akana) where he built core products for managing and monitoring web services.
January 21, 2016, 7-9 PM
(Networking 6:30-7:00 PM, and after the talk as time allows)
Simultantous Interpretation for Virtual Meetings
Hosted by Adobe
Photos: http://www.meetup.com/IMUG-Silicon-Valley/photos/26695954/
Slides/Video: see links in this article https://www.zipdx.info/...imug408-adobe-hq/
David Frankel, CEO and founder (left), and Barry Slaughter-Olsen, GM of Multilingual Operations
For the past four years, ZipDX has been developing a system to support simultaneous interpretation for “virtual meetings.” Conceptually, this is a multilingual conference call that behaves analogously to an in-person meeting conducted in a conference hall equipped with a multilingual sound system and interpreter booths. Participants connect using a conventional telephone and talk and listen in their preferred language, while interpreters use their magic to eliminate the language barrier. The interpreters connect to the audio conference via a web-based “virtual interpreter console.”
The solution, now in use by public and private organizations around the world, allows for “remote participation” (allowing remote participants to join an in-person meeting); “remote interpretation” (allowing interpreters to work remotely servicing an in-person meeting), and for fully-virtual meetings where interpreters and participants can all be in different locations.
Perhaps most interesting to language specialists are the discoveries the company has made as this solution has been deployed. These include issues with audio quality, lack of visual cues, coordination of work among and between interpreter teams, good and bad behaviors of meeting moderators and delegates, and treatment of presentation materials (such as PowerPoint slides). Medical and court interpreting, where the solution also applies, bring their own special considerations. The audio cues ZipDX has developed help participants with “situational awareness” during the meeting, and allow them to pace their speech to avoid overrunning the interpreter.
When things go well, the simultaneous interpreters eliminate the language barrier, and the result is a dynamic (almost musical) conversation in multiple languages.
David Frankel is the CEO of ZipDX, a provider of specialty virtual meeting services that he founded in 2007. The focus at ZipDX is to solve problems that cannot be addressed with conventional solutions, usually in “high-value” meeting situations that involve research, training, sensitive conversations, and participants speaking different languages. He has worked in high-performance computer and networking technology since 1974. He was the founder of Jetstream Communications, Inc. which supplied voice-over-packet technology to telecommunications carriers.
Get IMUG event announcements, sign up for reminders and RSVP via Meetup, the world's largest network of local groups. IMUG events are usually $5, but free for members, volunteers and hosts.
Join the multilingual computing community! Admission to most IMUG events is free for members. Membership is only $20, annual renewals are $15, and lifetime membership is $100.