Skip to content

About

Privacy-first linguistic humanizer & academic suite. 100% deterministic, rule-based, and zero generative AI.

Topics

Resources

Stars

2 stars

Watchers

0 watching

Forks

Latest commit

 

History

3 Commits

Folders and files

Repository files navigation

TurnHuman Community Edition Logo

⚡ TurnHuman Community Edition (TurnHuman CE)

Enterprise-Grade Linguistic Humanizer & Academic Suite
100% Programmatic, Deterministic Rule-Based, and Zero Generative AI.

License: GPL v3 Python: 3.10+ UI: Streamlit Privacy: In--Memory Compliance: GDPR & EU AI Act


🌟 Introduction

TurnHuman Community Edition (CE) is a fast, deterministic desktop linguistic engineering suite designed for researchers, academics, developers, and students. TurnHuman systematically transforms robotic, stiff, or machine-generated text into fluid, natural human prose while preserving scientific facts, technical jargon, formatting, and source code logic.

Unlike commercial tools that wrap third-party Large Language Models (LLMs), TurnHuman operates entirely via programmatic algorithms, Part-of-Speech (POS) tagging, syntactic dependency parsing, and mathematical n-gram disruption.

TurnHuman Community Edition Dashboard


💡 Why Choose TurnHuman CE?

Problem with Commercial LLM Humanizers The TurnHuman CE Guarantee
Factual Hallucination: LLM rewriters alter $p$-values, statistics, dates, and citations. Zero AI Hallucination: Deterministic rule-based transformations never fabricate or alter data.
Privacy & Leakage Risks: Documents are uploaded to cloud LLM servers and logged. 100% In-Memory Processing: Ephemeral RAM processing (io.BytesIO). Zero bytes saved to disk or cloud.
Secondary AI Footprints: LLMs produce flat burstiness profiles that trigger Turnitin AI & GPTZero. k-Gram Disruption: Mathematically dismantles 5-gram token chains and restores human burstiness.
Corrupted Documents: LLMs break Word XML styles, LaTeX formulas, and code syntax. Lossless In-Place Parsers: Complete preservation of .docx formatting, LaTeX math, and code logic.
Runaway Word Counts: AI rewrites expand texts, violating strict journal word budgets. Strict Zero-Expansion Mode: Guarantees exact word count preservation ($\Delta = 0$).

🧭 The 5 Academic Features & Tools

TurnHuman CE features 5 dedicated tools accessible via the modern interactive sidebar:

Feature & Tool Overview
1. Plagiarism & AI Humanizer Multi-stage linguistic pipeline for text and documents (.docx, .tex) with in-place formatting and zero word count expansion.
2. Code Comment Humanizer Naturalizes source code comments across multiple developer personas while keeping code logic 100% intact.
3. Citation & BibTeX Inserter Resolves verified peer-reviewed publications via live CrossRef DOIs into BibTeX, APA, and IEEE citations.
4. ZipPy Entropy Checker Compression-based entropy scanner that detects low-perplexity AI writing patterns.
5. Document Similarity Comparator Synchronized side-by-side diff viewer with Winnowing algorithms to highlight matching text segments.

📚 Supported Academic Taxonomies

TurnHuman CE incorporates decoupled, verified discipline-specific lexicons covering 5 broad academic faculties and 18 specialized subdisciplines:

  • Computer Science & AI: Artificial Intelligence & Machine Learning, Cybersecurity & Cryptography, Software Engineering, Systems & Networking.
  • Engineering & Physical Sciences: Fluid Dynamics & Thermodynamics, Signal Processing & Control Systems.
  • Mathematics & Statistics: Differential Equations & Dynamical Systems, Numerical Analysis & Optimization.
  • Medical & Life Sciences: Bioinformatics & Computational Biology, Genomics & Molecular Genetics, Public Health & Epidemiology, Pharmacy & Clinical Pharmacology, Nursing & Clinical Care.
  • Quantitative Finance & Economics: Option Pricing & Financial Engineering, Risk Modeling & Econometrics.

🚀 Quickstart & Installation

Prerequisites

  • Operating System: Windows 10/11, Linux (Ubuntu/Debian/Fedora), or macOS.
  • Python: Python 3.10, 3.11, 3.12, 3.13, or 3.14.

1. Clone or Download Repository

git clone https://github.com/andreiramani/turnhuman-ce.git
cd turnhuman-ce

2. Install Dependencies

pip install -r requirements.txt
python -m spacy download en_core_web_sm

3. Launch the Application

Option A: One-Click Launchers (Recommended)

  • Windows (Command Prompt): Double-click run.bat
  • Windows (PowerShell): Right-click and run with PowerShell: run.ps1
  • Linux & macOS: Run ./run.sh (or bash run.sh)

Option B: Terminal Command

streamlit run app.py

The application will launch and automatically open in your default browser at http://localhost:8501.


🔒 Privacy & Safety Assurance

  • Zero-Retention Guarantee: Your documents and text inputs are held strictly in ephemeral RAM (io.BytesIO) during processing. TurnHuman CE never writes user data to disk, temp folders, or databases.
  • Regulatory Compliance: Engineered in alignment with:
    • General Data Protection Regulation (GDPR) Article 5 (Data Minimization & Storage Limitation).
    • EU Artificial Intelligence Act (EU AI Act) Article 50 (Transparency and Ethical Human Oversight).
  • Anti-Steganography Sanitizer: Automatically cleans zero-width Unicode tracking markers (U+200B, U+200C, U+200D, U+FEFF, U+2060) and C2PA metadata embedded in generated texts.

  • Community Edition: Free for personal, academic, and open-source research use.

About

Privacy-first linguistic humanizer & academic suite. 100% deterministic, rule-based, and zero generative AI.

Topics

Resources

Stars

2 stars

Watchers

0 watching

Forks

Releases

Contributors

Languages