What Makes an Ideal Quote? Recommending "Unexpected yet Rational" Quotations via Novelty

Researchers developed NovelQR, an AI framework for recommending quotations that are 'unexpected yet rational' by prioritizing novelty over surface-level topical relevance. The system uses a generative label agent to interpret deep meanings and a novelty estimator to rerank candidates, showing superior performance in human evaluations across bilingual datasets.

• →Traditional quotation recommendation systems focus too heavily on topical relevance while ignoring semantic depth and aesthetic properties.
• →User studies reveal people consistently prefer quotations that are 'unexpected yet rational' in context, highlighting novelty as a key factor.
• →The NovelQR framework combines deep-meaning interpretation with token-level novelty estimation to improve recommendation quality.
• →Human judges rated the system's recommendations as more appropriate, novel, and engaging compared to existing baselines.
• →The approach addresses auto-regressive continuation bias that affects current AI models in understanding quotation meanings.