And finally we have metadata at Aftonbladet! Please don’t ask why we have waited until now :/
Metadata is the most important step Aftonbladet are taking this year as a product. We are done with the “Go Fish” approach . This is data infrastructure that allow us to do amazing stuff!
How are we thinking about metadata?
We have been very strict about how we speak about this in terms of terminology, I will explain why. Metadata is ”data about data” or data that describes the content. Example of that is publish date, section for the article and writer. That kind of MetaData is something we already have. But this does not describe the content very well. This is what we will add to get more information about the content.
Metadata Manager (MDM)
We are very serious about metadata and we have one person (so far) that is dedicated to work with the metadata at Aftonbladet. This is a very important role when we work with digital media. The MDM will be responsible for keeping the tag space and stories up to date and support the editors and reporters.
Tags
The tags or keywords describes the content, like where the story take place or who are in it. We have 4 different categories of tags to separate them and make it more understandable for the computers
- People – people or character
- Places
- Organisations – organizations, brands, sports teams etc.
- Descriptor – Fashion, Economy etc
We are very strict about the tags and it is very important to have a tag space that is accurate
Storyline
A storyline is basically a bunch of content that relates to the same thing. e.g. 9/11 in New York is a story.
Expected lifetime
We produce very much content every day and most of it has an expected lifetime of a day. And some of the content that we produce will be relevant for a week or longer. Most of the time the stuff we work on for a longer period will also be relevant for the users longer. In most cases we know how long it will be relevant for our users and we can use that. As a reporter or editor you can choose between:
- 1 Day
- 1 Week
- 1 Month
- Evergreen
- Specific date
This will enable the user to discover the most relevant content.
Character of the article
We will have even more automated lists of content presented to our users. Some personalized as I described here and some that is based on chronology. One of the things we put great pride in is the mix of content, some serious and some fun. This is the way to tell the computer if it is something tragic or fun.
What can we do with it?
As you might understand this gives us endless possibilities to present our brilliant content to the users.
- Users can follow a story or a person
- We can detect what kind of content a user reads. That give us the opportunity to do personalization
- We are going to find the articles!
- The Search engine optimization (SEO) is going to improve a lot!
- Video content can be found by google
How do we get this to work?
The most important thing about this is that everybody needs to put metadata on the content we create. If you do not have metadata on the content it is like sending a postcard without an address on it.
