Incremental Summarization using Taxonomy

In this paper, a new summarization system is proposed, which summarizes a document by interactively scoring the sentences using already-extracted summary so that the sen-tence which contains the most amount of relevant informa-tion to the summary to be extracted. To calculate the amount of relevant information contained by a sentence, the system makes heavy use of already existing taxonomies. The experiment shows a result pretty close to the current state-of-the-art systems