// the find
Sentdex/socialsentiment
Sentiment Analysis application created with Python and Dash, hosted at socialsentiment.net
A live-streaming Twitter sentiment analysis dashboard built with Plotly Dash and SQLite, created as a teaching companion for a Dash tutorial series. It streams tweets, stores them locally, and visualizes sentiment trends in real time. Primarily useful as a learning reference, not a production-ready tool.
The FTS (full-text search) SQLite approach for querying tens of millions of tweet records is a practical engineering choice that actually works at scale. The separation of streaming ingestion (twitter_stream.py) from the frontend (dash_mess.py) is the right architecture for a real-time pipeline. It's a complete end-to-end example — stream, store, query, visualize — which is rare in tutorial projects that usually stop at one layer.
Last pushed in 2018 and depends on the v1 Twitter Streaming API, which Twitter killed in 2023 — the core data pipeline is dead on arrival. The main app file is literally named dash_mess.py and the author acknowledges it's a mess; it's not a codebase to learn good patterns from. SQLite as the backend for millions of rows with concurrent writes from a streamer and reads from a web app is a reliability problem waiting to happen. No tests, no type hints, no documented schema.