What it does This is an awesome-list that catalogs sentiment analysis resources: survey papers, labeled datasets (AFINN, SentiWordNet, Stanford Sentiment Treebank), open-source implementations across seven languages, and SaaS APIs. It aims to serve two masters—academics hunting citations and developers hunting libraries.

The interesting bit The list explicitly flags a trap most gloss over: sentiment models are brittle across domains. A classifier trained on restaurant reviews may call your bug report “delicious.” The README embeds this warning twice, including a lobste.rs comment calling out naive off-the-shelf usage. That honesty is rarer than you’d think.

Key highlights

• Surveys the full stack from lexical methods (AFINN word lists) to attention-based deep learning (OpenAI’s byte mLSTM, memory networks for aspect-based analysis)
• Tracks the research-to-code lag: notes that aspect-level analysis dominates recent papers, but “not all open-source implementations are caught up yet”
• Language coverage is broad if uneven—Python gets 10 entries, C# gets one Naive Bayes classifier
• Includes a single web app (Textalytic) and six commercial APIs for the “just pay someone” crowd

Caveats

• Last meaningful update appears to be several years ago; the OpenAI byte mLSTM link and several paper URLs may rot
• Coverage depth varies wildly by language; Ruby and C# entries are thin
• No benchmarking or comparison data—it’s a phone book, not a Consumer Reports

Verdict Worth bookmarking if you’re entering sentiment analysis and need a map of the territory. Skip it if you already know your way around Hugging Face and want performance numbers; this won’t tell you which library wins on your dataset.