Section Metadata Reference

---

Decision‑Making Insights

• Content pacing: Monitor words, paragraphs, and read_time_min to balance chapter lengths.
• Tone analysis: Use paragraph_sentiments to spot narrative highs and lows within a chapter.
• Keyword strategy: Leverage tf_idf and topics to refine SEO, marketing tags, or blurb keywords.
• Style adjustments: Compare readability and lexical_metrics against target audience grade levels.
• Entity tracking: Review entities to ensure consistent character/place naming throughout sections. ``

How each analytics tile can guide your edits

UI tile / metric block	What it reveals	How to use it to polish the current chapter
Core counts (words, paragraphs, read_time_min)	Pacing & density of the chapter.	If this chapter is 8 000 words while others average 3 000, consider splitting or tightening. A 40-min read may exhaust some readers.
Readability (Flesch–Kincaid, Fog, SMOG)	Sentence complexity & estimated U.S. grade level. • Flesch–Kincaid = grade level based on sentence length and syllables. • Gunning Fog = grade level based on sentence length and complex words. • SMOG = grade level based solely on polysyllabic words.	Aim for 4–6 for middle-grade, 6–9 for YA, and 10+ only for academic or literary prose. Lower scores mean more accessible reading. Adjust by simplifying sentence structure and vocabulary.
Lexical metrics (ttr, hapax, avg_word_length)	Vocabulary breadth & repetition.	Low TTR (< 0.25) signals echoing words—add synonyms or sensory details. Very high TTR (> 0.45) can read purple; prune unusual adjectives.
Bag of Words & Phrases (term_freqs, bigrams, tf_idf)	Over-used lexis and signature phrases.	If “wolf growled” appears 15 ×, try “snarled”, “bared its teeth”, or non-verbal cues to keep imagery fresh. Promote rare, vivid TF-IDF terms into hook lines or chapter titles.
Sentiment gauge (sentiment)	Net emotional tone of the whole section.	A strongly negative compound (e.g., –0.99) may need a breather scene or light banter before the next down-beat.
Paragraph sentiment arc (paragraph_sentiments)	Moment-to-moment emotional flow.	Spot flatlines: seven neutral paragraphs in a row? Inject a surprise, revelation, or sensory jolt to maintain momentum.
Sentence metrics (avg_len, min_len, max_len)	Rhythmic texture of prose.	Avg 12 words with max 45 suggests many short beats punctuated by an occasional ramble—merge tiny sentences for smoother cadence.
Entities (add-on)	Character & place consistency.	Catch misspelled names (“T’y” vs “Ty”) or verify that minor locations don’t vanish without closure.
Topics / Genre (add-on)	Dominant themes & market fit.	If “political intrigue” topics outweigh “romance” in what you intended as a love story, refocus scenes or marketing copy.

Interpretation (section-level example)

• Words 7 420 → longest chapter so far (≈25 min read); split or tighten for steadier pacing.
• F-K 10.2 → dense senior-HS level; shorten multi-clause sentences to ease flow.
• TTR 0.21 → vocabulary slipping into repetition; replace echoes like repeated “looked” with “glanced, surveyed, scanned.”
• “back” × 32 (top term) → filler word; swap for spatial cues (“stepped away from the doorway”).
• “wolf growled” × 12 (top bigram) → image losing impact; vary with “snarled,” “lips curled,” “hackles bristled.”
• Sentiment –0.88 → emotionally heavy throughout; add a hopeful beat (joke, sunrise) to avoid reader fatigue.
• Dialog 62 % → dialogue-driven scene; weave in sensory details between lines to ground setting.

Interpretation — Section-level Paragraph Sentiment Trend

• Paragraph polarity bars
  Each stacked bar mirrors the emotional ratio inside one paragraph. Sudden flips (red→green) flag mood pivots—great for reveals.

• Compound line
  Paragraph-by-paragraph single score.
  • A steep drop (+40 → –60) marks shock moments.
  • If the line flatlines near zero, tension may stall—inject conflict or humor.
  • Remember: bars give the mix; compound gives the net.

Revision workflow:

1. Smooth big word-count outliers to even pacing.
2. Shape document-level sentiment for balanced highs/lows (use compound & bars together).
3. Fine-tune micro-beats with paragraph trend inside each scene.

info

Advanced analytics—Named-Entity checks, topic clustering, trope/beat detection, and AI engagement scoring—are optional add-ons at checkout for deeper structural and marketing insight.

Use this checklist hierarchy: fix readability and high-frequency fillers first, adjust emotional pacing second, then leverage advanced add-ons for final polish.

Aggregated insights at the section level

Field	Type	Description
title	Option<String>	First heading detected (H1/H2/H3).
subtitle	Option<String>	Secondary heading (e.g. H2 following H1).
author	Option<String>	Author attribution extracted from a “By …” line in the opening paragraphs.
words	usize	Total word count in this section.
paragraphs	usize	Number of <p> tags/paragraphs.
chars	usize	Total character count, including spaces.
read_time_min	usize	Estimated reading time (ceil division at 200 WPM).
term_freqs	HashMap<String, usize>	Top N word frequencies (stop words removed).
bigrams	HashMap<String, usize>	Top N two-word phrases.
tf_idf	HashMap<String, f64>	Top N TF–IDF keyword scores relative to the full document.
sentiment	Option<Sentiment>	Overall sentiment scores (neg/neu/pos/compound).
paragraph_sentiments	Vec<Sentiment>	Sentiment score for each paragraph—useful for sentiment arcs.
entities	Option<Vec<(String,String)>>	Named-entity tokens and labels (e.g. PERSON, ORG).
topics	Vec<String>	Extracted topics via LDA or similar modeling (stub for future integration).
genre	Option<String>	Heuristic genre label based on top keywords.
sentence_metrics	SentenceMetrics	Avg/min/max sentence length (in words).
lexical_metrics	LexicalMetrics	TTR, hapax legomena count, average word length.
readability	ReadabilityMetrics	Flesch–Kincaid, Gunning Fog, SMOG scores.
dialog_ratio	f64	Proportion of this section’s words that occur within detected dialogue passages.