README: KNEXT Output from the Brown Corpus
Jonathan Gordon, 2011-12-07

---

This contains the results of running a development (not release) copy of
KNEXT on the Brown Corpus:
  Kucera, Henry & W. N. Francis. 1967. Computational Analysis of
  Present-Day American English. Brown University Press.

The logical form is documented in the files accompanying the KNEXT
source[1]. Beyond that, note that when predicates end in '*' it is to
indicate that they have been abstracted from a more specific form.

These results have been lightly filtered to exclude propositions that
are likely to be overly vague, too specific, or malformed (e.g., missing
arguments).

[1]: http://www.cs.rochester.edu/research/knext/download

---

brown.kb:
  This is a merged knowledge base, where each line is of the format
    <count> <proposition> <verbalization>
  The count is the number of times this proposition was found in the
  Brown corpus. The proposition is KNEXT's logical form. The
  verbalization is an English-like representation of the formula that
  can be used, e.g., when asking human judges to rate the knowledge.

