Analyze the sentiment of Hong Kong market news.
The news are scrapped from AASTOCKS, and the sentiment is determined by the users' (likely retail investors) rating on the articles.
Several models will be used and evaluated to see which one is the best.
This project is still in progress
Chapter 1: Scrapping news from AASTOCKS
Chapter 2: Loading and visualizing data
Chapter 3: Lexicon based approaches
Chapter 4: Unigrams and n-grams, and resampling the training data
Chapter 5: spaCy