The LTi News Roundup - 25th June 2012 (part 2)

Weekly news round-up prepared by the Editorial Staff of LangTechNews for LT-Innovate, the Forum for Europe’s Language Technology Industry.



Voices of/for Customers

We all know about the fashionable concept of ‘voice of the customer’. But how about inventing a ‘voice of the supplier’ or even the physically voiceless? European speech innovators, faced with massive multilinguality and acutely attuned to demographics and dialect, are leading the field in synthesising voices to boost speech interaction. Acapela Group has joined forces with AssistiveWare to provide synthetic children’s voices for those who cannot speak, integrating them into an app for use on phones. At the same time, CereProc has extended its voice portfolio with two new voices of different genders and accents to developers who need to build more personalised voice applications. Like Acapela, CereProc works in the assistive field to help disability, but then builds on this experience to open up its technology to a broader range of applications. As well as the expanding voice market on mobile devices, the e-book phenomenon offers a fast-growing market for text-to-speech applications.

Leveraging the language long tail

Google last week announced funding for a new project (covered in Common Sense Advisory’s Global Watchtower blog) dedicated to saving endangered languages. The idea is to record, collect and manage audio and video data of speakers of these tongues. It forms part of a more global semi-coordinated effort between academia, private foundations and other sponsors to act quickly to preserve this critical component of the human heritage. In a slightly different move, the translation technology facilitator TAUS is taking the lead in ways of generating more translations between as many human languages as possible, including the resource-poor languages that Google will be supporting. One solution is the pivot language strategy that creates new data resources by linking two languages via a common intermediary. Google’s collection effort should be able to help TAUS’ production model by using recorded (and transcribed) data as the basis for training a translation engine. And then swarming the bilingual translation process across many pairs of languages.

Market Byte

Analysts say that UK, Spain, France, Germany, and Italy have turned into nations of digital shopkeepers. They accounted for 60% of the European online retail market last year. But Russia is set to grab 30% of this market by 2016, according to companiesandmarkets.com

The LTi News Roundup - 25th June 2012 (part 1)

Weekly news round-up prepared by the Editorial Staff of LangTechNews for LT-Innovate, the Forum for Europe’s Language Technology Industry.

Top of the Week: Naturalising the interface. 

After a classic career path from technology start-up to acquisition by an industry major, Swype (a Nuance company since last year) delivered an interesting Beta upgrade this week to its input technology combining speech technology, clever word statistics and keyboarding.
As the population of mobile users explodes, the interface between user and device is becoming ever more critical. Mobile voice assistants made their debut towards the end of 2011 and the strategic battle for quality and language coverage is on-going. Gesture and touch are now being experimented with or manufactured into a variety of personal and collective (e.g. TV) screens.
This latest Swype product, then, packs swyping (predictive text), voicing (speech input), touching, and typing in a single app. It learns from your emails to predict your typed-in words, and also teaches the voice interface (based on Dragon technology) to better prepare for your typical voice input. This takes us one step further down the road towards personalized devices offering more choice to the end-user, and greater leverage of cloud-based machine learning to automate language capabilities,

New Ways to Smarten up Content: There are two basic entry points to ‘intelligent content’

One is access (what smarts will help me find and use existing content of all kinds?), the other is creation (how can I improve my new content so that it integrates as relevantly as possible into global conversations?).
On the creation dimension, the European Commission’s own Joint Research Centre this week announced the availability of its EuroVoc Indexer that automates the categorisation of parliamentary documents in 22 languages. This makes life easier for content curators in large institutions, who do not have to index documents manually: the machine learns what the document is about so that it is then easier to access.
Similarly, the Dutch start-up Silk has developed an application that automatically prepares semantic content for websites. It adds in the relevant metadata and open-data links that turn a website’s content into ready to go “semantic web data” that can be exploited by other semantic tools in this growing content space. A good example of how complex technology can be packaged into user-friendly solutions for everyday publishers of digital content.
Along the access dimension, the Barcelona company Cognicor has innovated with a new application for automating the processing of low-value complaints. Customer complaints form a genre of user-generated content that is soaring in volume as commercial enterprises of all kinds open up their websites to customer feedback. Cognicor’s idea is to relieve busy contact centre agents by automating the entire process of understanding a complaint, checking its validity and even negotiating some form of compensation. This means using semantics in various interesting ways to interpret the language of the complaint.

The LTi News Roundup - 4th May 2012 (part 2)

Weekly news round-up prepared by the Editorial Staff of LangTechNews for LT-Innovate, the Forum for Europe’s Language Technology Industry.

What's happening in the Language Technology industry, from LT-Innovate


Intelligent content:

The Polish company Positionly has netted another round of funding to develop its SEO analytics solution that informs companies large and small about their positioning on web search results without needing to know anything about the topic. A great example of how language-sensitive functionality can make self-knowledge easier for busy companies. The venerable Italian lang tech developer SynTHEMA has netted a contract with the Italia newspaper La Stampa to automatically categorise editorial content using semantic technology. And the Irish start-up SindiceTech is helping companies derive real business value (and not just technical advance) from the linkage power of semantic web data.

Thanks to Manx: 

A special mention this week for the exemplary efforts of the Isle of Man - one of Europe’s smallest countries (80K population) and one of the world’s oldest continuous parliamentary bodies. The Manx parliament has digitised and published online a large collection of its records, making them easy to search and treat as a data resource for anything from linguists to lawyers. And it has been experimenting with speech recognition technology (developed partly by the UK firm VoicePower) to ensure that its current parliamentary records are captured live as a speech stream as well as being converted into text (and edited) to speed up the publication process.

Speaking of Market Size: 

Common Sense Advisory published its latest projection for 2012 on the market for outsourced translation and interpretation services from some 21,000 suppliers, sized at $33.5B, growing by 12.17% annually. The language technology portion of this sum was not communicated about directly, and the figure does not of course include the size of the in-company translation workload. Meanwhile the global speech recognition is set to market to grow by 15.65% over the next three or so years according to another market research firm. Which suggests some kind of uniform growth rate in these very different sectors. The fact is that the world’s translation activity as a whole (either paid for or free) is a vast below-the-radar market, often unrelated to technology. We therefore have little useful knowledge of its true size. Speech recognition on the other hand is exclusively mediated through technology and much easier to map. But perhaps the most fascinating data this week came from YouTube, with a claim that 4B videos are viewed every day, 70% of them from outside of the U.S. A trove of speech genres, images, and languages just waiting to become a useful resource, unless Google monetises it first!

The LTi News Roundup - 4th May 2012 (part 1)

Weekly news round-up prepared by the Editorial Staff of LangTechNews for LT-Innovate, the Forum for Europe’s Language Technology Industry.

What's happening in the Language Technology industry, from LT-Innovate

Top of the Week: 

Swiftkey, the UK text input company, has brought out a new (beta) version of its predictive text input app for mobiles. X Keyboard is now an established member of the Android ecosystem with millions of downloads and plenty of paid-for added-value versions into the bargain. The secret ingredient is Swiftkey’s next word prediction technology. This keeps it ahead of the next-letter input models used by competing solutions, and won it the Most Innovative Mobile App at Global Mobile Awards in Barcelona earlier in 2012. By building up huge databases of word usage patterns, the company will probably be able to extend X Keyboard to specific user domains such as tablet-toting healthcare workers and similar enterprise input situations. The big question, of course, is whether speech input will ultimately win out over text input in the vast mobile device marketplace, and make virtual keyboards as old-hat as mouse ‘n keyboard hardware. Both techniques will most likely co-exist for some time.

The Language Learning Galaxy: 

The German company Speexx was awarded the 2012 Golden Engine Award for Strategically and Sustainably Driving China’s Education Industry last week, for the best contribution to improving the country’s education industry with a number of cloud-driven European language solutions. There are a number of signs that language learning will be impacted by the striking changes underway in the global learning marketplace. The publisher Pearson has just acquired GlobalEnglish to boost its position in the English language learning market, and the UK gaming app player Mindshape has been given funding to develop an innovative language learning game to be called Language City. It looks as if language learning aids can be adapted to any user platform, mobile or desktop, and benefit by leveraging data from users’ errors and successes through a cloud architecture. Just as the digital education agenda as a whole will morph towards College as a Service, language learning will need to take on more language processing smarts to customise courses and lessons (as we used to call them) more intelligently to individual profiles and needs.

The LTi News Roundup - 13th th May 2012 (part 2)

Weekly news round-up prepared by the Editorial Staff of LangTechNews for LT-Innovate, the Forum for Europe’s Language Technology Industry.

What's happening in the Language Technology industry, from LT-Innovate




Source-language engineering : 

The Berlin-based information quality company Acrolinx helps contentcentric organisations write documents that are easier to understand, translate and search. Gartner has now added their name to its Cool Vendors A-list. Making text ‘translatable’ used to be most relevant to exporters with mission-critical products that simply could not afford to get Lost in Translation. Today, all website content is a potential export product, and intelligent writing – i.e. writing for reading/translating/SEO – should head the list of everyone’s content strategy agenda. A nice example this week came from the Czech localisation company Moravia’s blog on reworking English words into a Japanese order to ready text for machine translation. This sounds complicated because it requires lots of upstream editorial work. But it is likely that rewriting (or pre-editing) source text could be (semi-)automated, as was predicted over 50 years ago at the very first MT conference. Nobody likes the term ‘controlled language’ any longer, even though tweeting is a form of it. Think of source engineering as a long-term investment: if you want to machine-translate language pairs that include German or Japanese (two languages that computers find particularly hard to translate into), building up corpora of quality parallel texts by introducing a source-fixer today will pay off a thousand-fold tomorrow.

xHealth :

The healthcare sector is growing prefixes – mHealth, eHealth – making it an increasingly interesting target for language technology applications. Indeed “intelligent healthcare” – iHealth – is a major theme in LT-Innovate activities. IDC believes we are entering a second wave of mobile healthcare.
Laptops are the dominant mobile device today (84% of installed base), but smartphones and tablets will eventually dominate the market and spending on mobile point-of-care solutions in Western Europe will nearly double by 2015. In Europe, big healthcare data is fragmented across countries and languages, making it hard for applications to gain critical mass, as the Publicis Healthcare Communication Group has shown. But as the recent eHealth Week in Copenhagen demonstrated, individual countries are successfully transforming health data into digital intelligence. Transcription of audio recordings, speech interfaces, language interpretation for patients, ontologies and text analytics are all playing a critical role in enabling medicine to evolve into iHealth.

The LTi News Roundup - 13th th May 2012 (part 1)

Weekly news round-up prepared by the Editorial Staff of LangTechNews for LT-Innovate, the Forum for Europe’s Language Technology Industry.

What's happening in the Language Technology industry, from LT-Innovate

Top of the Week : Exalead

Calling the French firm Exalead a ‘search company’ is beginning to sound silly. Now part of the Dassault Systèmes stable of product lifecycle management companies, it is stepping up its role as all-round innovation accelerator. This week it announced that two French start-ups will be developing new applications based on public data as part of Exalead’s role as incubator for the government Dataconnexions program. The idea is for these two young companies to explore marketable uses of government data, and one will be working on health data and the other on hyper-local information. They will both leverage Exalead’s Cloudview technology to develop their apps.

The Semantic Revolution

‘Semantics’ can be a weasel word in the ITC sector. In the discipline of linguistics it partners phonology and syntax as a level of language structure that addresses the meaning of words and sentences. In industry, on the other hand, semantics refers to the capacity to link data items together (ideally tagged with additional information).This can produce insights and applications that represent more than the sum of their parts. The German company Transinsight scored twice this week with its inventive data search-and-show technology driven by semantics. One project will be delivering an automated RSS news analyst to the Mayor of Dresden’s office which will enable rapid evaluation and decision-making. In a completely different field, it will provide interactive modelling software for the Norwegian company Statoil to help them visualise geodata. Other semantics companies in view include the UK firm Ontology Systems that ensures enterprise data alignment, and Concept Searching which released a new version of its flagship product.