Macropodus

Why we included this project

Macropodus covers most of the Chinese text-processing chores a small team would otherwise assemble from several separate libraries. In one package you get word segmentation, part-of-speech tagging, named entity recognition, keyword extraction, text summarization, and text similarity, with the core segmentation and tagging running on an Albert+BiLSTM+CRF model trained on large Chinese corpora. It also includes less common helpers: new-word discovery, a scientific calculator that parses arithmetic written in natural language, Chinese-to-Arabic and Roman numeral conversion, pinyin conversion, and simplified/traditional character conversion. For a Chinese-language app that needs preprocessing plus a few text-analysis utilities, that makes it a convenient single dependency. The function-based API drops into existing code easily, though some model files are downloaded separately after installation.