Trike - machine learning based content analysis and prediction system (Round 6)

Share:
spacer
Finland Project type: large Helsingin Sanomat; Sanoma Media Finland Oy; Sanoma Group Oyj

Summary

Engaging content increases likelihood of a reader subscribing to news media. Trike will be a tool to help journalists write more engaging content, faster, with the help of data. The tool will analyse content and customers’ engagement with the content, utilise the data to enable journalists to make more data-optimised content decisions, and give real-time feedback to journalists during the writing process.

The solution

The current content authoring process does not utilize user engagement data to its full potential to make decisions. The lack of article level engagement data also restricts our ability to optimise algorithmic content delivery to reflect actual content engagement. This solution builds upon creating a data collection and analysis process to record customer engagement with article content. With the data set, Helsingin Sanomat will build a machine learning-based system to deliver the engagement data and predictions to journalists during writing.