# MLSNT: Multi-Lingual Social Network Toxicity Dataset
**MLSNT** is a multi-lingual toxicity classification dataset created via LLM-assisted label transfer, designed to support scalable, resource-efficient moderation in gaming and social media contexts.
This dataset accompanies the paper:
> **Unified Game Moderation: Soft-Prompting and LLM-Assisted Label Transfer for Resource-Efficient Toxicity Detection**
> Accepted at **KDD 2025**, Applied Data Science Track
---
## 🧩 Overview
MLSNT standardizes toxicity annotations across **7 languages** and **15 open-source datasets** using GPT-4o-mini. It enables:
- Binary and fine-grained toxicity classification
- Cross-lingual training with consistent labels
- Production-ready deployment
---
## 🌍 Supported Languages
- **French** (`fr`)
- **German** (`de`)
- **Portuguese (Brazilian)** (`pt`)
- **Russian** (`ru`)
- **Simplified Chinese** (`zh-cn`)
- **Traditional Chinese** (`zh-tw`)
- **Japanese** (`ja`)
---
## 🏗️ Construction Metho…
