Bootstrappausten

Bootstrappausten is a fictional term used in discussions of computational stylistics and semi-supervised text generation to describe a bootstrapping-based approach to mastering a specific author's style, notably Jane Austen's. The name combines bootstrap (self-starting learning) with Austen, signaling an iterative process that grows its knowledge from a small starting corpus and expands through self-generated data.

Concept and process: A small, high-quality Austen-like corpus is used to train an initial model. The model then generates new candidate texts in the target style. A filtering stage—based on stylistic metrics, coherence, or human evaluation—selects high-confidence outputs. These become additional training data, and the model is retrained in cycles, gradually improving its ability to imitate the author's voice.

Applications: educational tools for teaching stylistic analysis, creative writing aids, or research in stylometry and authorial imitation.

Variants: used also for other authors or genres.

Limitations: risk of overfitting to surface features; copyright and authenticity concerns; potential for abuse in deception.

See also: semi-supervised learning, bootstrap aggregating, style transfer, generative authorship.