Eigenword Resource Page

An Eigenword is an real-valued vector "embedding" associated with a word that captures its meaning in the sense that distributionally similar words have similar eigenwords. This page contains links to several sets of eigenwords They are computed as the singular vectors of the matrix of co-occurrence of words and their contexts, and used in a variety of spectral NLP methods and applications. See our NAACL Tutorial on spectral methods

Eigenword Collections

eigenwords for many languages and code for computing them

Note that the "words" include punctuation, and so will confuse some software.

Coming soon: eigencontexts

Software for plotting eigenwords

matlab/octave plotting code
- plot_eigenwords.m which calls pick_words_csv.py and read_word_name.m
R/python plotting code
- plot_eigenwords.R which calls pick_words.py

home: ungar@cis.upenn.edu