Sunday, November 29, 2015

online psycholinguistics demos 2015

I was asked recently about an old post from 2008 that listed a variety of online psycholinguistics demos. All of the links are dead now, so I was asked if I knew of any updated ones. This is what I can find. Any suggestions would be welcomed.

  • Harvard Implicit Associations TaskProject Implicit is a non-profit organization and international collaboration between researchers who are interested in implicit social cognition - thoughts and feelings outside of conscious awareness and control. The goal of the organization is to educate the public about hidden biases and to provide a “virtual laboratory” for collecting data on the Internet.
  • webspr - Conduct psycholinguistic experiments (e.g. self-paced reading and speeded acceptability judgment tasks) remotely using a web interface
  • Games With Words: Learn about language and about yourself while advancing cutting-edge science. How good is your language sense?
  • Lexical Decision Task demo: In a lexical decision task (LDT), a participant needs to make a decision about whether combinations of letters are words or not. For example, when you see the word "GIRL", you respond "yes, this is a real English word", but when you see the letters "XLFFE" you respond "No, this is not a real English word".
  • Categorical PerceptionCategorical perception means that a change in some variable along a continuum is perceived, not as gradual but as instances of discrete categories. The test presented here is a classical demonstration of categorical perception for a certain type of speech-like stimuli.
Paul Warren has a variety of demos at the site for his textbook "Introducing Psycholinguistics"

  • McGurk demo

  • Various other demos from Warren's textbook

  • Saturday, November 14, 2015

    Google's TensorFlow and "mathematical tricks"

    TensorFlow is a new open source software library for machine learning distributed by Google. In some ways, this could be seen as a competitor to BlueMix (though much less user friendly). Erik Mueller, who worked on the original Watson Jeopardy system (and has a vested interest in AI with his new company Symbolic AI), just wrote a brief review of TensorFLow for Wired.

    Google’s TensorFlow Alone Will Not Revolutionize AI

    Unfortunately, it's not really a review of TensorFlow itself, but rather makes a general point against statistical approaches, which I basically agree with, but the argument requires a much more comprehensive treatment.

    Some good quotes from the article:

    • "I think [TensorFlow] will focus our attention on experimenting with mathematical tricks, rather than on understanding human thought processes."
    • "I’d rather see us design AI systems that are understandable and communicative."

    Nuts and Bolts of Applying Deep Learning (Andrew Ng)

    I recently watched Andrew Ng's excellent lecture from 2016 Nuts and Bolts of Applying Deep Learning and took notes. I post them as a he...