View a full list of current and prior NULab events:http://web.northeastern.edu/nulab/events/events-all/
- This event has passed.
Ben Schmidt: Using word2vec and other vector space models to explore relationships in language
November 9, 2016 @ 1:30 pm - 3:30 pm
In recent years, Google’s word2vec and other vector space models of language have emerged as simultaneously useful and intriguing ways to computationally describe human language as a set of relationships among words. They have proven considerably better than previous methods at tasks like describing language analogically. (In the canonical example, for instance, word2vec can accurately describe that a queen is the female equivalent of a king.)
In this workshop, we’ll learn to interpret and create word2vec models using R. We’ll cover training models on your text: looking for similarities and relationships; and branch out into some of the major open questions particularly interesting for those in the humanities and computational social sciences, such as mathematically removing things like gender bias from a data set or comparing different models to each other.
This event is free and open to the public, but if you are not a member of the Northeastern community, please email Sarah Connell at sa.connell[at]neu[dot]edu to arrange access to the library.