The University of Texas at Austin
  • Science & Technology

    When Will My Computer Understand Me?

    By Aaron Dubrow, Texas Advanced Computing Center
    Published: Aug. 13, 2013

    The difference between “charging” a battery and “charging” a client is easy for people to understand. But for a computer, distinguishing between the various meanings of a word can be problematic. Linguistics researcher Katrin Erk has developed a new way to help computers learn natural language.

    For more than 50 years, linguists and computer scientists have tried to get computers to understand human language by programming semantics as software. Driven by efforts to translate Russian texts during the Cold War (and more recently by the value of information retrieval and data analysis tools), these efforts have met with mixed success. IBM’s Jeopardy-winning Watson system and Google Translate are high-profile, successful applications of language technologies, but the humorous answers and mistranslations they sometimes produce are evidence of the problem.

    chart illustrating how words with similar meanings cluster together

    A “charge” can be a criminal charge, an accusation, a battery charge or a person in your care. Some of those meanings are closer together, others futher apart.

    Our ability to distinguish between multiple word meanings is rooted in a lifetime of experience. The context in which a word is used — an intrinsic understanding of syntax and logic, and a sense of the speaker’s intention — all help us interpret what another person is saying. A computer can’t access these experiences, so it requires a lot of data to begin to “learn” the distinctions.

    But language isn’t always straightforward, even for humans. The multiple definitions in a dictionary can make it difficult even for people to choose the correct meaning of a word. Katrin Erk, a linguistics researcher in the College of Liberal Arts, refers to this as “semantic muck.” Enabled by supercomputers at the Texas Advanced Computing Center, Erk has developed a new method for visualizing the words in a high-dimensional space.

    Instead of hard-coding human logic or deciphering dictionaries to try to teach computers language, Erk decided to try a different tactic: feed computers a vast body of texts (which are a reflection of human knowledge) and use the implicit connections between the words to create a map of relationships.

    “An intuition for me was that you could visualize the different meanings of a word as points in space,” Erk says. “You could think of them as sometimes far apart, like a battery charge and criminal charges, and sometimes close together, like criminal charges and accusations (“the newspaper published charges…”). The meaning of a word in a particular context is a point in this space. Then we don’t have to say how many senses a word has. Instead we say: ‘This use of the word is close to this usage in another sentence, but far away from the third use.’ ”

    To create a model that can accurately recreate the intuitive ability to distinguish word meaning requires a lot of text and a lot of analytical horsepower.

    “The lower end for this kind of a research is a text collection of 100 million words,” she explains. “If you can give me a few billion words, I’d be much happier. But how can we process all of that information? That’s where supercomputers come in.”

    Watch the video to learn how Erk uses the technology at the Texas Advanced Computing Center to help computers “learn” natural language.


    An earlier version of this story appears on the TACC website.

    • Quote 2
      John said on Aug. 27, 2013 at 10:21 a.m.
      I also agree with Matt. You really can't teach a computer anything and they are only going to be as "smart" as the thought process of the human being that is programming them. I have a Google Android phone and it irritates me to no end when the phone tries to second-guess what I am typing and corrects me as I go. It has caused many errors in posts to friends on Facebook and other social media. Artificial intelligence is only as intelligent as the person doing the programming.
    • Quote 2
      Shana Lott said on Aug. 25, 2013 at 5:15 p.m.
      While I don't foresee singularity, That day is not too far when there will be no perceivable difference between humans and computers as far as analytical understanding goes.
    • Quote 2
      GM Lim said on Aug. 25, 2013 at 7:50 a.m.
      Agree with Matt. Having said that it looks simple to me. We are the creature of the lord of mighty God while computers are not. Will be Hard to understand human languages until we-human understand how the lord created human being.
    • Quote 2
      Matt said on Aug. 21, 2013 at 10:02 p.m.
      Semantic muck indeed! To be fair, computers don't 'understand' anything, they are simply machines that do what they are programmed to do. The first step is for humans to understand what computers really are. They are nothing more than abstraction processing machines which have not the ability to "understand" the abstractions they process but only to process abstractions as they are programmed to do. Artificial Intelligence is artificial by definition and the appearance of intelligence in computers is nothing more than an active image of human thought processes captured and put into the stone of computer hardware to process. So to increase the "appearance" of intelligence we only need to capture more human thought processes and map them in a manner that is accessible.. Of course the way to do this is to recognize the functions we humans cannot avoid the use of and program the computer to have this functionality, that we may be better able to capture and map images of human mental processing in a manner of machine processing ability. Interesting topic nonetheless!
    • Quote 2
      Ducktoes Computer Services said on Aug. 20, 2013 at 7:49 p.m.
      Hopefully this happens soon, as it will make life way easier!
    • Quote 2
      Active Computing said on Aug. 19, 2013 at 3:47 a.m.
      Once computers can understand human speech, I believe the world will change. Ordering food at a restaurant seems like it would be one of the first things to change.
    • Quote 2
      khoa cua said on Aug. 16, 2013 at 9:52 p.m.
      This is advancement of linguistics in technology is fascinating.Great article!
    • Quote 2
      Brett said on Aug. 16, 2013 at 10:53 a.m.
      This is advancement of linguistics in technology is fascinating. Although interacting with "Siri" on my iPhone and with other verbal interactive software still have a ways to go to achieve effective verbal communication between computer and human, we have really come a long ways in recent years. Great article!
    Share:
    • Digg
    • del.icio.us
    • StumbleUpon
    • Facebook
    • Google Bookmarks
    • LinkedIn
    • Twitter
    • Print
    • email

    Related Topics

    , , , ,

  • Top Videos

    • ...
    #HiddenUT: There’s More to UT Than You Know
    #HiddenUT: There’s More to UT Than You Know
    From secret nooks to unexpected resources, untold collections, artifacts and treasures...
    10 Things You Need to Know About UT System’s Next Chancellor
    10 Things You Need to Know About UT System’s Next Chancellor
    Longhorn and Navy SEAL Admiral William McRaven was named the sole finalist to become...
    Around the World Tour: Longhorns Play Asia, London
    Around the World Tour: Longhorns Play Asia, London
    Summertime is world tour season for big-time musicians: Bruno Mars, Katy Perry and...
    Tech and Health: Innovative Devices Aren’t Science Fiction
    Tech and Health: Innovative Devices Aren’t Science Fiction
    From nanomotors to customized prosthetics, check out inventions developed by Cockrell...
    Lifelike Heart Valve Model Will Aid Treatment
    Lifelike Heart Valve Model Will Aid Treatment
    Michael Sacks is developing mitral valve models that will enable heart surgeons to...
    (1:00)
    Dell Med Kickoff Invites Community Input
    Dell Med Kickoff Invites Community Input
    Construction kicked off for the Dell Med School at a ceremony Monday. Leaders, including...
    How Do You Reinvent Health Care?
    How Do You Reinvent Health Care?
    Start by re-imagining medical education, says Dell Med School Dean Clay Johnston....
    Preventing Earthquake Damage
    Preventing Earthquake Damage
    A team of Cockrell School engineers is working in New Zealand to understand soil...