Title :
New feature sets for summarization by sentence extraction
Author :
van Halteren, Hans
Author_Institution :
Nijmegen Univ., Netherlands
Abstract :
Machine learning features developed for authorship attribution can also be used for sentence extraction. This is done in order to achieve the automatic production of summaries based on key sentences. The author´s pilot extraction system outperformed a hard-to-beat positional baseline system.
Keywords :
abstracting; learning (artificial intelligence); natural languages; authorship attribution; feature sets; key sentences; machine learning features; positional baseline system; sentence extraction; summarization; summary; Data mining; Feature extraction; Information retrieval; Machine learning; Vocabulary; Writing;
Journal_Title :
Intelligent Systems, IEEE
DOI :
10.1109/MIS.2003.1217626
