Rich interfaces for reading news on the web
Proceedings of the 14th international conference on Intelligent user interfaces
| Hi-index | 0.00 |
Software that models how types of news events unfold can extract information about specific events and explain them to a news reader. This support can be useful when the background provided by an article is insufficient, if other news coverage exists from which an event's history can be extracted. For extended sequences of related events, it is reasonable to expect that articles published after the sequence concludes include less background coverage of the sequence. Focusing on two stereotypical types of event sequences --- kidnappings and corporate acquisitions -- we distinguish between articles providing in-depth coverage, those having multiple sentences mentioning the same event sequence, from articles making a passing mention in just one sentence. We find that, after an event sequence concludes, passing mentions become more common and there are significantly fewer mean mentions per article.