Predicting User Views in Online News

Daniel Hardt, Owen Rambow

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

97 Downloads (Pure)

Abstract

We analyze user viewing behavior on anonline news site. We collect data from64,000 news articles, and use text featuresto predict frequency of user views.We compare predictiveness of the headlineand “teaser” (viewed before clicking) andthe body (viewed after clicking). Both arepredictive of clicking behavior, with thefull article text being most predictive.
Original languageEnglish
Title of host publicationProceedings of the 2017 EMNLP Workshop: Natural Language Processing meets Journalism
EditorsOctavian Popescu, Carlo Strapparava
Number of pages6
Place of PublicationStroudsburg, PA
PublisherAssociation for Computational Linguistics
Publication date2017
Pages7-12
ISBN (Electronic)9781945626883
Publication statusPublished - 2017
EventNatural Language Processing Meets Journalism: EMNLP 2017 Workshop - København, Denmark
Duration: 7 Sept 20177 Sept 2017

Workshop

WorkshopNatural Language Processing Meets Journalism
Country/TerritoryDenmark
CityKøbenhavn
Period07/09/201707/09/2017

Cite this