Orellana-Rodriguez, ClaudiaClaudiaOrellana-RodriguezGreene, DerekDerekGreeneKeane, Mark T.Mark T.Keane2017-04-182017-04-182016 ACM2016-05-259781450342087http://hdl.handle.net/10197/8427ACM Web Science 2016 (WebSci ’16), Hannover, Germany, 22-25 May 2016News media face many serious concerns as their distribution channels are gradually being taken over by third parties (e.g., people sharing news on Twitter and Facebook, and GoogleNews acting as a news aggregator). If traditional media is to survive at all, it needs to develop innovative strategies around these channels, to maximize audience engagement with the news they provide. In this paper, we focus on the issue of developing one such strategy for spreading news on Twitter. Using a corpus of 1M tweets from 200 journalist Twitter accounts and audience responses to these tweets, we develop predictive models to identify the features of both journalists and news tweets that impact audience attention. These analyses reveal that different combinations of features influence audience engagement differentially from one news category to the next (e.g., sport versus business). From these findings, we propose a set of guidelines for journalists, designed to maximize engagement with the news they tweet. Finally, we discuss how such analyses can inform innovative dissemination strategies in digital media.en© ACM, 2016. This is the author's version of the work. It is posted here by permission of ACM for your personal use. Not for redistribution. The definitive version was published in WebSci '16: Proceedings of the 8th ACM Conference on Web Science http://doi.acm.org/10.1145/2908131.2908154.Machine learningStatisticsComputational journalismSocial mediaAudience engagementNews eventsConference Publication10711610.1145/2908131.29081542017-01-13https://creativecommons.org/licenses/by-nc-nd/3.0/ie/