Metadata-Version: 2.4
Name: budgetnlp
Version: 0.0.3
Summary: NLP on a budget
License: MIT
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: flashtext
Dynamic: description
Dynamic: description-content-type
Dynamic: license
Dynamic: license-file
Dynamic: requires-dist
Dynamic: summary

# budgetnlp

This is a library to do "good enough" nlp. It is intended to be used with SEC data via the [datamule](https://github.com/john-friedman/datamule-python) library. If this ends up being useful for something else, neat!

Supports two types of data:
1. Text
2. Data Tuples. Unique data type used by the [doc2dict](https://github.com/john-friedman/doc2dict) library.

Focus:
- "good enough" quality.
- cheap.

This is designed with SEC documents (legal) in mind. Legal documents have certain structural constraints that allow shortcuts to be taken.

This library will likely start out rough, as I am not yet an expert on ML.

