TSSuBERT: How to Sum Up Multiple Years of Reading in a Few Tweets.

The article focuses on leveraging pre-trained language models like BERT to enhance the summarization of tweet streams, introducing an extractive model that adapts the summary size automatically. It also presents a novel methodology for constructing tweet stream summarization datasets with minimal human effort and releases a new dataset called TES 2012-2016, potentially advancing research in incremental summarization.