Personalized summarization of scientific scholarly texts

Alka Khurana, Vasudha Bhatnagar, Vikas Kumar

Abstract: Highlights•Propose an unsupervised algorithm for generating extractive personal summaries.•Handle negative and positive knowledge preferences of the user.•Leverage wNMF to uncover latent semantic space.•Introduce a multi-granular framework for personal summary evaluation.•Evaluate on four public data-sets containing scientific articles.