Learn TALL,
step by step.

The interface, the Settings panel, the buttons you will meet on every page, and the ideas behind the vector space model and text pre-processing.

• Interface overview13
• Settings14
• TALL buttons15
• Pre-processing steps17

Four ways to build a corpus — text files, Biblioshiny exports, saved .tall projects, Wikipedia — then split, sample and enrich it with metadata.

• Import texts18
• From Biblioshiny19
• Load .tall files20
• Sample collections21
• Edit24

Filter and group documents by their metadata, assign time, label and keyness roles, and choose the parts of speech that enter the analyses.

• Filter36
• Groups37
• Feature roles38
• PoS tag selection39

Corpus size and lexical metrics, vocabulary and TF-IDF, wordclouds, frequencies by PoS, morphological features and the dependency tree viewer.

• Main information41
• Vocabulary & TF-IDF42
• Wordcloud43
• Morphological features45
• Dependency tree viewer46

Syntactic complexity, topic models (LDA, CTM, STM), supervised classification, polarity and emotion detection, extractive and abstractive summaries.

• Syntactic complexity81
• Topic modeling85
• CTM & STM93
• Supervised classification96
• Polarity detection99
• Emotion analysis105
• Extractive summarization109
• Abstractive summarization113

TALL AI interprets plots and tables with Google Gemini; the report collects the results of every page into a single Excel workbook.

• TALL AI115
• Report116
• Authors & references117