Slanguage is a web application designed to bridge the language gap across generations and cultures by translating slang, memes, and modern expressions into universally understandable language.

As Gen Z and Gen Alpha continue to shape their own digital dialects through trends and online culture, Slanguage helps users stay connected and communicate clearly — no matter their age or background.

🔗 Demo Video: Watch here

1. Clone the repository

   git clone https://github.com/yourusername/Slanguage.git
   cd Slanguage

2. Install dependencies

   npm install

3. Set up environment variables
   Create a .env file and add your configuration settings (e.g., API keys, MongoDB credentials).
4. Run the app

    npm run dev

5. Open your browser and go to your local host

Hackers: Cris, Irene, Wilson, Natasha

Hackathon: Hack the Valley X — University of Toronto Scarborough

💬 Feedback and contributions are welcome!

Open an issue or submit a pull request to help improve Slanguage.

1. Collect multi-platform contexts

   cd backend
   npm run collect:data -- --collectors reddit,youtube \
     --reddit-subs slang,teenagers --reddit-limit 200 --reddit-comments true \
     --youtube-query "slang explained" --youtube-comments true

   Configure REDDIT_*, YOUTUBE_API_KEY, and friends in backend/.env before running.

2. Run the C++ analyzer for insights

   cd backend/src/training/cpp
   g++ -std=c++17 -O2 slang_trainer.cpp -o slang_trainer
   ./slang_trainer --input ../../data/generated/slang.contexts.tsv \
     --output ../../data/generated/slang_language_model.json \
     --top-tokens 20 --related-limit 8 \
     --graph-output ../../data/generated/slang_related.tsv \
     --state-out ../../data/generated/slang_stats.dat \
     --clusters 10 --embedding-features 48 --min-pmi 0.05

   Next reloads can resume from the saved state with --state-in. Output now includes PMI-weighted context tokens, per-phrase quality scores, k-means cluster assignments, and a TSV edge list for graph/cluster tooling.