The Usage of GPLWordNet

Introduction

GPL WordNet (Generating Polarity Lexicions with WordNet) is a lexicon-generating tool that uses WordNet as a database to find two sets of lexicons.

It provides two basic function,

• Full expansion: This function is used to find target words' antonyms, synsets, adjectives, derivational words, and hyponyms.
• Antonyms expansion - This function is used for finding target words' antonyms.

This tool is an extended Python version of Nicolas et al.'s dictionary-generating approach. As their team only provides the R version, it is quite hard to use. Therefore, I provide this Python version, and you can absolutely improve and self-define this package's function by git my code from my GitHub.

Function

1. example.full_expand()
2. example.antonyms_expand()

Agruments

Input Parameters

1. full_expand() function

Notes: Only key 'term' is compulsory.

2. antonyms_expand() function

Notes: Only key 'term' is compulsory.

Output Parameters

1. full_expand() function

2. antonyms_expand() function

Example

1. full_expand() function

Basic function

data = [{'term': 'good'}]
print(example.full_expand(data))

data = [{'term': 'good', 'PoS': 'ADJECTIVE'}]
print(example.full_expand(data))

Extra Parameters

data = [{'term': 'good', 'PoS': 'ADJECTIVE'}]
print(example.full_expand(data,syns=True,antonym=False))

2.antonyms_expand() function

Basic function

data = [{'term': 'good', 'PoS': 'ADJECTIVE'}]
print(example.antonyms_expand(data))

Extra Parameters

data = [{'term': 'good', 'PoS': 'ADJECTIVE'}]
print(example.antonyms_expand(data,syns=True))

Supplemantal Materials

1. This work is an replication work of Nicolas' SADCAT codes, which only have R version. For more details about Nicolas Works, please refer to following link,

   • Nicolas dictonaries Webpage.
   • Nicolas GitHub.

2. The approach I mainly learned from Maks et al.'s(2014) Paper.

   • Generating Polarity Lexicons with WordNet propagation in five languages.

Call for Contributions

The GPLWordNet project welcomes your expertise and enthusiasm!

Small improvments of fixes are always appreciated, please submit your feature according GitHUb.

Our preferred channels of communication are all public, but if you’d like to speak to us in private first, contact my public email at xuanlongqin.cu@gmail.com .

Acknowledge

I would like to sincerely thank my dear Ling Yan, who spent three hours accompanying me and watching TV series alone during this period, supporting me to complete this project.