linguistics
Latest
Why humans can't use natural language processing to speak with the animals
We’ve already got machine-learning and NLP that can translate speech into any number of languages. Surely adapting that process to animal calls shouldn't be that big of a stretch?
A simple string of 'and's seems to crash Google Docs pages
Why "And. And. And. And. And." breaks Google Documents pages is a mystery, but it does.
Magnetic retainer lets you operate machinery with your tongue, gives linguistics new meaning
Many people with debilitating conditions rely on sip and puff technology to get around, which is relatively lo-fi by some standards. Now, a special retainer with magnetic sensors could bring mobility into the smartphone age. Developed at Georgia Tech, the Tongue Drive System uses a magnetic piercing to track lingual gestures. The sensors then transmit data to an iOS app that translates it to on-screen or a joystick movement. Earlier versions used a headset, but the prototype revealed at the International Solid State Circuits Conference in San Francisco, is hoped to be more comfortable and discreet. The system is currently being trialled by 11 participants with high-level spinal-cord injuries, with larger trials planned.
NTT DoCoMo exhibits on-the-fly speech translation, lets both parties just talk (video)
The race to smash linguistic barriers with simultaneous speech-to-speech translation is still wide open, and Japanese mobile operator NTT DoCoMo has just joined Google Translate and DARPA on the track. Whereas Google Translate's Conversation Mode was a turn-based affair when it was demoed back in January, requiring each party to pause awkwardly between exchanges, NTT DoCoMo's approach seems a lot more natural. It isn't based on new technology as such, but brings together a range of existing cloud-based services that recognize your words, translate them and then synthesize new speech in the other language -- hopefully all before your cross-cultural buddy gets bored and hangs up. As you'll see in the video after the break, this speed comes with the sacrifice of accuracy and it will need a lot of work after it's trialled later in the year. But hey, combine NTT DoCoMo's system with a Telenoid robot or kiss transmission device and you can always underline your meaning physically.
Microsoft keeps gunning after Apple's 'generic' App Store trademark, brings in a linguistics expert
We'd say this was getting silly but that would imply that it wasn't already. Microsoft and Apple are still at each other's throat over the latter's trademark application for the term "App Store," with Microsoft now bringing in a Dr. Ronald Butters, Professor Emeritus at Duke University and a man with a taste for hardcore semantics. He says the compound noun "app store" is perfectly generic in that it "does not merely describe the thing named, it is the thing named." In a wildly geeky turn, he references the potential for someone discovering a use for masers and trying to trademark the term "maser store" in response, which would seem immediately and logically absurd. An app store, says the good doctor, is no more capable of being trademarked than a grocery store or a stationery store or a computer store. Of course, as with most trademark disputes, what's truly at stake here isn't linguistics, but a big fat wad of consumer goodwill. Having previously been quite uncomfortable with the idea of buying additional software for his mobile phone, Joe Consumer has nowadays grown quite accustomed to dropping little chunks of change on smartphone apps, and the terminology that sets his mind at ease most readily is indeed "app store." Preventing others from using that well established moniker would clearly be a significant competitive advantage for Apple and it's pretty hard to argue with its contention that it's responsible for generating the goodwill that sits behind it. Then again, we reckon Android's Market, webOS' admittedly small App Catalog, and other moves by the likes of RIM, Nokia and Microsoft itself with WP7, haven't done the app store cause any harm either, so in purely ethical terms it still seems a little rich for Apple to be claiming the app store crown all to itself. As to the legal battle itself, it's descending into quite amusing minutiae, but its outcome will be of great interest to most of the aforementioned mobile ecosystem purveyors.
Expanding the Dwarvish tongue in Lord of the Rings Online
Some things are just going to appeal to a certain group of players, and that's fine. If you're enjoying Lord of the Rings Online for game design, class balance, or the like, then you might not be terribly concerned with the nuances of the functional languages that Tolkien developed for the original books. But if you really care about giving your Dwarf weapons with more resonant names than "Foe-hewer," you probably would be interested in A Casual Stroll to Mordor's latest post about expanding the rather anemic Dwarven tongue. Using the Hebrew linguistic structure as a base, the article goes into depth about making logical alterations to existing words, and from there, creating new words that fit into the established structure. It's a fascinating post for Lord of the Rings Online players as well as anyone who likes developed in-game lore, as it shows just how much can be done within the given setting. And if you're a player in LotRO who'd like to use some more Dwarven names, there are interesting charts available along with some example names.
New MIT software learns an entire dead language in just a few hours
Whenever we boot up our time machines, cruise back to 1200 B.C., and try to pick up chicks at our favorite wine bar in Western Syria, our rudimentary knowledge of Ugaritic is usually more embarrassing than helpful. The good folks at the Massachusetts Institute of Technology have us stoked on some new software we hope to have in pocket form soon. It analyzes an unknown language by comparing letter and word patterns to another known language (in Ugaritic's case, its close cousin is Hebrew) and spits out a translation quickly, using precious little computing power. To give some perspective, it took archaeologists four years to do the same thing back in 1928. It's not quite Berlitz yet, but this proof of concept is kind of like the Michael Jordan of computational linguists -- it's probably the first time that machine translations of dead scripts has been proven effective. If we plug some hopeful numbers into our TI-83, we calculate that we'll be inserting our own genes into the ancient Syrian pool in a matter of months. Thanks, MIT! [Photo courtesy of Wikipedia Commons]
IBM's Watson is really smart, will try to prove it on Jeopardy! this fall (video)
As much as we love our Google homepage, computer search remains a pretty rudimentary affair. You punch in keywords and you get only indirect answers in the form of relevant web results. IBM doesn't seem to be too happy with this situation and has been working for the past three years on perfecting its Watson supercomputer: an array of server racks that's been endowed with linguistic algorithms allowing it to not only recognize oddly phrased or implicative questions, but to answer them in kind, with direct and accurate responses. Stuffed with encyclopedic knowledge of the world around it, it answers on the basis of information stored within its data banks, though obviously you won't be able to tap into it any time soon for help with your homework. The latest word is that Watson's lab tests have impressed the producers of Jeopardy! enough to have the bot participate in a televised episode of the show. That could happen as early as this fall, which fits right in line with our scheduled doom at robots' hands by the end of 2012. Ah well, might as well get our popcorn and enjoy the show.
