Aller Media

AI Content Assister

Aller Media is a modern and innovative media house that delivers content and entertainment to readers and users throughout the Nordics, digitally as well as in print, every month.

Challenge

Aller Media needed to develop an automated, digital solution that could add relevant metadata to their existing image database.

Metadata can e.g. be identification of people, description of the image’s surroundings, etc.

The purpose was to ensure an increased use of the existing image database and at the same time improve the user experience by providing access to more metadata from a better and more detailed tagging of the individual images.

Solution

The developed solution handles all images that Aller has in their image database across all of their four Nordic markets.

The developed solution, Aller AI Content Assister (AICA), is built on top of AWS Rekognition – a standard AI tool from AWS for person recognition in images.

Aller AICA has also been expanded with an interface to Woodwing’s Elvis digital asset management platform, which is already used at Aller.

In addition, a user interface has been developed which is designed for handling deviations and allocating access rights. The user interface is designed so that it is possible to create a country-specific user access, so that users are only presented with relevant images. The solution also includes a fully flexible dashboard, which displays the system’s key data.

Result

The project of developing AICA was completed in two months from the completion of the joint design document to the completion of the first phase. The first phase has today been implemented in all Aller’s Nordic markets.

Aller estimates that the potential savings from the implementation of Aller AICA are around EUR 3.70 per picture. This estimate does not include the potential quality improvements from better use of existing images resulting from more detailed metadata on the images and resulting more frequent use of existing images.