Language Tech Market News
52.6K views | +21 today
Follow
Language Tech Market News
The Home of Multilingual Intelligence
Curated by LT-Innovate
Your new post is loading...
Your new post is loading...
Scooped by LT-Innovate
Scoop.it!

UK-based @Volume Releases Tool for Marketers to Train #NLProc Engines

UK-based @Volume Releases Tool for Marketers to Train #NLProc Engines | Language Tech Market News | Scoop.it

UK-based marketing agency Volume, specialized in conversational solutions, has launched the “first Training and Testing-as-a-Service platform for natural language data models.”
Called QBox, it helps marketers test whether the training data used to fine-tune a NLP engine is working as it should. QBox, CEO Chris Sykes said in a statement, “allows natural-language data model developers to look into the black box they are typically working with, to easily understand any impact of a change on their training data.” 

LT-Innovate's insight:

Services developing around conversational tech...

more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

Open-Sourcing Data Licence Proposed by Linux

Open-Sourcing Data Licence Proposed by Linux | Language Tech Market News | Scoop.it

The Linux Foundation has created a new open license framework, Community Data License Agreement (CDLA), which may do for data what open source did for programming. At Open Source Summit Europe, The Linux Foundation announced a new family of open-data licenses. The CDLA licenses are an effort to define a licensing framework to support collaborative communities built around curating and sharing "open" data.

more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

Time to Build a Language Data Market? - TAUS

There is naturally much debate about how much data a given neural system needs to build a baseline engine that can drive a translation production line for a given semantic domain. Responses depend on multiple parameters and issues: a lot of data for a baseline, much less than before for a domain-specific? Experimentation is all. But both quality and quantity are key parameters: the real question is, how can the availability of language data in general be optimized to enable better translation?
more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

@lingvist: Datafication for Language Learning

@lingvist: Datafication for Language Learning | Language Tech Market News | Scoop.it

Estonian startup Lingvist aims to help us pick up a new language faster than usual.  Founder Mait Müntel was working as a physicist at CERN, and wanted a quick way to learn French. So he decided to take a data-driven approach, and fed over 40,000 movies complete with French subtitles into an algorithm that was running on the CERN supercomputers.  Through this, he was able to identify the key words and phrases that make up the bulk of French speaking, which in turn enabled him to master around 95% of those words in just a few months. This approach has been used successfully by around 1 million users who are testing and refining the system as well as learning one of a handful of languages via the app.  In initial tests, Müntel claims that learners have outperformed peers undertaking traditional classroom based tuition by four times.

LT-Innovate's insight:

Data needs context to be useful for human learning. How did he program in context/situation to his usage data?

more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

UK Gov-Backed Report Calls for AI 'Data Trusts'

UK Gov-Backed Report Calls for AI 'Data Trusts' | Language Tech Market News | Scoop.it

UK government report on AI: Among the 18 main recommendations by the experts is that the UK Government, AI companies and private businesses should work together to create Data Trusts, i.e. legally binding, long term framework arrangements that share valuable data with AI companies to help them to create new applications. ‘To use data for AI in a specific area, data holders and users currently come together, on a case by case basis, to agree terms that meet their mutual needs and interests,’ the report authors explain.

LT-Innovate's insight:

Another way around the data accessibility dilemma

more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

Massive Data Sets are Not a Commodity for AI

Massive Data Sets are Not a Commodity for AI | Language Tech Market News | Scoop.it

There is no open market for training data. I suspect there never will be because many organizations closely guard data as premium among their intellectual property. Let’s stick with our autonomous driving example. Companies in this industry are in a race to get to Level 4 autonomy, where cars drive on their own. It’s not likely the automakers will share their proprietary data in the midst of competition this fierce. Nor will banks, insurance providers, e-commerce merchants, advertisers, or, given the choice, many of us.

LT-Innovate's insight:

Suggests there will not be a digital data marketplace in the near future. But could "language data" as used in translation and speech (AI) applications constitute a special legal case of shareable data? Their value would be in the algorithms and networks, not just in the "real-world content" of the data.

more...
No comment yet.