Content Spinner™
Semantic Equivalence-Controlled Text Variation Engine
Content Spinner™ generates semantically equivalent but lexically distinct text variations. Unlike naive word-swap spinners, it uses synonym databases, syntactic restructuring, phrase-level alternatives, and semantic similarity verification to produce unique variants that read naturally and preserve the full meaning of the original content.
How It Works
Spintax Generation
Builds a spintax tree for the input text — a structured representation where alternatives are provided for words, phrases, and sentences at multiple granularity levels. Spintax format: {word1|word2|word3} allows deterministic or random variant selection.
Synonym Candidate Scoring
For each candidate synonym, scores: semantic similarity (embedding cosine), register match (formal/informal), frequency match (common word preference), and collocation validity. Only synonyms scoring > 0.75 composite are included in spintax.
Structural Spinning
Beyond word-level spinning, applies sentence-level structural spin: reordering coordinated clauses, moving adverbials, converting between active/passive constructions, and substituting equivalent sentence patterns (e.g., "X is Y" → "Y characterizes X").
Uniqueness Verification
After generating each variant, measures overlap with other variants using normalized edit distance and n-gram similarity. Ensures each generated variant has < 30% surface-level overlap with any other variant.
Key Metrics
Use Cases
- A/B testing different content versions for conversion optimization
- Multi-channel publishing with unique content per platform
- Generating unique product description variants for e-commerce
- Creating test data sets with controlled content variation
Try Content Spinner™ Now
Powered by our proprietary technology stack. Available directly in the TextHumanize tools and via REST API.