Week 13 · Stylistic Features and Vectorisation 文体特征识别与向量化

Dec 7, 2026 · ← Course home

Planned session. From text to numbers. We identify the features that carry style at scale — function words, part-of-speech distributions, n-grams, sentence length, type-token ratio, readability indices — and learn to turn a folder of texts into a feature table: one row per text, one column per measurement. This “vectorisation” step is the bridge between Week 4’s lexical intuitions and Week 14’s models; suddenly, a style is a point in space.

Zero-code tooling throughout: taggers and spreadsheet workflows, with Python as an optional deep-dive for the curious.

Key concepts: feature engineering; function words as style fingerprints; n-grams; vector representation of texts.

Slides and the feature-extraction handout will be posted here after class.