stylometric-ai-detector 0.2.2

AI vs Human text detection using stylometric features and Random Forest

stylometric-ai-detector

AI vs Human text detection using stylometric features and a Random Forest classifier.

Overview

This library provides two functions:

• extract_stylometric_features(text) — Extract 8 stylometric features from any text.
• predict(text) — Classify text as AI-generated or human-written, with a confidence score.

The model is a Random Forest classifier achieving 96% accuracy, trained on the AI vs Human Text dataset. On first use, the model is automatically downloaded from Hugging Face and cached locally.

Installation

pip install stylometric-ai-detector

Quick Start

from stylometric_ai_detector import extract_stylometric_features, predict

# Extract stylometric features
features = extract_stylometric_features("The quick brown fox jumps over the lazy dog.")
print(features)
# {'char_count': 44, 'word_count': 9, 'avg_word_len': 4.0, 'punct_count': 1, 'sentence_count': 1, 'avg_sentence_len': 9.0, 'upper_case_count': 0, 'title_case_count': 1}

# Predict AI vs Human (model auto-downloaded from Hugging Face on first call)
result = predict(text="Artificial intelligence is transforming our world.")
print(result)
# {'label': 'AI', 'probability': 0.99}

# Or pass pre-computed features
result = predict(features=features)
print(result)
# {'label': 'AI', 'probability': 0.91}

Stylometric Features

Feature	Description
char_count	Total number of characters
word_count	Total number of words
avg_word_len	Average word length
punct_count	Number of punctuation characters
sentence_count	Number of sentences
avg_sentence_len	Average sentence length (in words)
upper_case_count	Number of fully uppercase alphabetic words
title_case_count	Number of title-case words

Model

The trained Random Forest model is hosted on Hugging Face at dinisds/stylometric-ai-detector. It is downloaded and cached to ~/.cache/stylometric-ai-detector/ on first use — no extra setup needed.

Dataset

Trained on Shanegerami's AI vs Human Text dataset from Kaggle. The dataset contains ~487k text samples labeled as human-written (0) or AI-generated (1).

License

MIT