pun-nlp 0.0.5

A robust NLP pipeline for stemming, lemmatization, and vectorization

NLPProcessor

Overview

NLPProcessor is an automated, adaptive NLP pipeline that dynamically handles:

• Tokenization (Word & Sentence)
• Stopword Removal
• POS Tagging
• Named Entity Recognition (NER)
• Text Normalization (Lowercasing, Punctuation Removal, etc.)
• Stemming & Lemmatization (via NLTK or spaCy)
• Vectorization (TF-IDF or Count Vectorizer)
• Dependency Management (Auto-installs missing libraries)
• Support for 2D Text Arrays (Processes lists of lists of text)
• Exception-Free Execution (Handles API changes without breaking)

Features

• Automated dependency installation
• Works with both NLTK and spaCy
• Vectorization support using scikit-learn
• Handles single strings and 2D arrays
• No human intervention required

Installation

Run the following command to install missing dependencies:

pip install pun_nlp

Usage

Import and Initialize

from pun_nlp import NLPProcessor

processor = NLPProcessor(stem=True, lemmatize=True, vectorize="tfidf", backend="spacy")

Process a Single Text

output = processor.process("running jumped swimming")
print(output)

Process a 2D Array of Text

input_texts = [
    ["I am running", "He is jumping"],
    ["They are swimming", "Dogs are barking"]
]
output = processor.process(input_texts)
print(output)

Customization Options

Parameter	Description
stem	Enable stemming (default: False)
lemmatize	Enable lemmatization (default: False)
vectorize	Choose "tfidf", "count", or None (default: None)
tokenize	Enable word/sentence tokenization (default: False)
remove_stopwords	Remove stopwords (default: False)
pos_tagging	Enable Part-of-Speech tagging (default: False)
ner	Enable Named Entity Recognition (default: False)
normalize	Lowercase and remove punctuation (default: False)
backend	Choose "nltk" or "spacy" (default: "nltk")

Check Supported Vectorizers

print(NLPProcessor.supported_vectorizers())  # ['tfidf', 'count']

License

This project is licensed under the MIT License - see the LICENSE file for details.