Grammar Rule Extractor

Domain : Natural Language Processing

Sub-Domain : Language Processing

Techniques : Grammer Rule Extraction

Application Domain : Language Modeling

Description

1. Extracted from the BROWN file all grammar rules embedded in parse trees. (Did not consider punctuation as a nonterminal and eliminated numbers attached to non-terminals such as '-1', '-2', etc.)
2. Reported the following:
   1. number of distinct rules found.
   2. the 10 most frequent rules regardless of the non-terminal on the left-hand side
   3. the non-terminal with the most alternate rules (i.e. the non-terminal that can have most diverse structures).
3. Estimated the above grammar size if lexicalized (i.e. to add head words to some of the rules).

Languages : Python

Tools/IDE : Anaconda

Libraries :

Duration :

Current Version : v1.0.0.13

Last Update : 03.16.2018 (Time : 04:20 P.M)