murray-z/text_analysis_tools

Leverages pre-trained word embeddings for synonym and semantic similarity tasks, requiring users to supply their own embeddings. Provides a modular API structure with included test datasets and executable examples for rapid prototyping. Combines classical NLP techniques (keyword extraction, text clustering) with neural approaches for tasks like text correction and event triple extraction from Chinese text.

733 stars. No commits in the last 6 months.