Online newspaper

Going online created more opportunities for newspapers, such as competing with broadcast journalism in presenting breaking news in a more timely manner.

Austra Taylor, author of the popular book, The Peoples Platform, argues that online news does not provide the detail needed to fully understand what actually happened.

Newspapers with specialized audiences such as The Wall Street Journal and The Chronicle of Higher Education successfully charge subscription fees.

The Guardian experimented with new media in 2005, offering a free twelve-part weekly podcast series by Ricky Gervais.

An example is the UK Southport Reporter, introduced in 2000—a weekly regional newspaper that is not produced or run in any format than 'soft-copy' on the Internet by its publishers, PCBT Photography.

Unlike blog sites and other news websites, it is run as a newspaper and is recognized by media groups such as the NUJ and/or the IFJ.

allNovaScotia is an online newspaper based in Halifax, Nova Scotia, Canada that publishes business and political news six days a week.

[13] In the US, technology news websites such as CNET, TechCrunch, and ZDNet started as web publications and enjoy comparable readership to the conventional newspapers.

Also, with the ever-rising popularity of online media, veteran publications like the U.S. News & World Report are abandoning print and going online-only.

[18] There are some newspapers which are predominantly online, but also provide limited hard copy publishing[11] An example is annarbor.com, which replaced the Ann Arbor News in the summer of 2009.

[16] The turn to hybrid publishing models has been commensurate with the increasing importance of social media platforms to disseminate news, especially amongst 18-24 demographic.

This is in line with the Pew Research Center's[22] finding in a survey of U.S. Americans that the Internet is a leading source of news for people younger than 50.

The number of times an article gets shared on social media is relevant for activists, politicians, authors, online-publishers and advertisers.

With new methods of Natural Language Processing such as Latent Dirichlet allocation it is possible to gain insights into the core characteristics of an article.

The features consist of variables describing words, links, digital media, time, keywords, insights from Natural Language Processing and the number of article shares.

30 November 2020. used machine learning methods, namely, logistic regression, linear discriminant analysis, artificial neural networks and random forests to predict the top ten percent most frequently shared articles.