HUANGZHIHAO1994/weibo-analysis-and-visualization

Implements a modular preprocessing pipeline with MongoDB JSON document ingestion, supporting both simplified and traditional Chinese text normalization via zh_wiki conversion. Architecture chains specialized preprocessing scripts (prepro.py, pre_graph.py, senti_pre.py) with downstream analysis modules, enabling flexible workflows from raw Weibo scraper output to multiple visualization outputs. Integrates word2vec embeddings with LSTM neural networks for sentiment classification and supports both dictionary-based and deep learning approaches to text analysis.

949 stars. No commits in the last 6 months.