Singapore Press Holdings has agreed a partnership with public sector body The Agency for Science, Technology & Research (A*Star) to use its AI technology to optimise SPH’s headlines.
In a bid to improve digital audience engagement, A*Star’s Institute of High Performance Computing will use its proprietary sentiment and emotion analysis technology to predict the virality of an SPH article based on its headline.
SPH journalists can then test different headlines for their articles before selecting the one most effective at driving engagement. SPH said past studies show that up to 80% of readers merely scan headlines instead of reading full articles, and that the headline is one of the most important factors in deciding whether to read an article.
While online publishers already put out different variations of headlines for the same article to test and analyse reader response, SPH said this use of AI will go deeper by allowing the media brand to identify specific components of a headline that resonate with readers and influence their reading decisions.
“This project will help our newsrooms better understand the emotional impact of different words and phrases used in headlines, so we can improve our engagement with readers and guide them towards important and informative news without resorting to clickbait,” said Anthony Tan, SPH deputy CEO.
The agreement builds on SPH’s efforts to incorporate new technology into its digital platform, and follows its partnership with DC Frontiers to use machine learning to develop a content recommendation engine.
Meanwhile, SPH has also launched its own digital stock photo marketplace, Photonico. The library draws on SPH’s photo archive, dating back to 1948, and offers a range of images depicting Asian culture, heritage and contemporary events. Photonico will be available to users for a fee.