Keynotes | K&Ts | GACs | Talks | Posters | Search

Poster C in Poster Session C: Wednesday, August 5, 9:30 – 11:15 am, Kimmel Center, Shorin & Rosenthal Rooms

A Simple Biologically Plausible Neural Model is Capable of Basic Natural Language Aqcuisition

Daniel Mitropolsky¹, Christos H. Papadimitriou²; ¹Massachusetts Institute of Technology, ²Columbia University

Presenter: Daniel Mitropolsky

Despite tremendous progress in neuroscience, we do not have a compelling narrative for the precise way whereby the spiking of neurons in our brain results in high-level cognitive phenomena such as reasoning and language. NEMO is a simple formulation of six basic and broadly accepted principles of neuroscience: excitatory neurons, brain areas, random synapses, Hebbian plasticity, local inhibition, and inter-area inhibition. We implement with NEMO a simulated neuromorphic system capable of basic language acquisition: starting from a tabula rasa, the system learns, in any language, the semantics of words, their syntactic role (verb vs. noun), the word order of the language, and the ability to generate novel sentences, all through the exposure to a number of grounded sentences that is similar to language acquisition by humans. The difficulty of learning the word order of the language decreases with the prevalence of the word order among world languages.

Topic Area: Auditory, Speech & Language Processing