Systematic evaluation and benchmarking of text summarization methods for biomedical literature: From word-frequency methods to language models

The rapid expansion of biomedical literature demands automated summarization tools that reliably condense research articles into concise, accurate summaries.

The rapid expansion of biomedical literature demands automated summarization tools that reliably condense research articles into concise, accurate summaries. We benchmarked 62 summarization methods, ranging from frequency-based and TextRank extractors to encoder-decoder models…