Web annotation is already at scale

Three takeaways from I Annotate 2015


by An Xiao Mina, Ed Bice and Chris Blow

What does web annotation look like at scale? This was one of the prompts at I Annotate 2015, a conference focused on, well, web annotation. Annotation is critical to Meedan’s view of the world: we believe that annotation is fundamentally an act of contextualizing content, and meaning travels better to other cultures and languages with the addition of contextualization.

So, we wanted to bring our Meedan use cases and thinking to the proceedings to see what’s what. Here are some of our takeaways:

Annotation comes from the human need to talk about things.

What is the impetus behind annotation? During a philosophical conversation in a breakout group, we tried to define annotation. Ed asserted that an annotation is any thing capable of containing meaning that can be differentiated from and attached to a source.

After a good deal of debate about what an annotation can be and what it looks like, the group did come to the consensus that anything can be annotated. Thus, annotation is motivated by the human need to talk and explain; in this way, annotation provides a structure for extending or depicting the meaning of things in the world.

Annotation is already at scale. Now we need better methods to make sense of them.

As Maryann Martone of Hypothes.is noted in a group presentation, annotation is already at scale. Indeed, a lot of media items receive likes, retweets and shares, which could be considered annotations of value to the reader. Thomas Richardson, aka the “Anticonsultant,” posted a great photo of the number of places where we can find annotations. It’s clear, though, that these annotations aren’t necessarily interoperable, and there’s no easy way to aggregate them into one place.

Importantly, we’ll also need to make better use of moderation and filtering tools to help make sense of the number of annotations being generated across different communities and contexts. I was particularly struck by one group’s concept of a configurable “filter” that could help make sense of the many annotations across the web, bringing value to different groups of people.

It’s an exciting time to build standards and interoperability so annotations can communicate with each other. As we work on our own products, we’re starting to tackle these questions: for instance, how can our verification work on Checkdesk interact with the idea of translation as an annotation that we’re building into Bridge?

We need to develop more specific use cases and frameworks that inspire people to participate.

Rap Genius, Climate Feedback and Fold all show great specific use cases for annotation. Annotation at scale is a noble goal, but showing real world applications will be critical to embedding the practice around the web.

Genius, formerly known as Rap Genius, famously started around people’s passion for annotating and elucidating the meaning behind rap lyrics. Climate Feedback uses the Hypothes.is platform so experts in climate science can verify and debunk claims around the web. Fold takes a pretty interesting graphical approach to annotation, by embedding and showing live media alongside source text.

A screenshot from Genius showing annotations of Bone Thugs N Harmony’s “Crossroads”.

While it’s important to figure out the technical details of implementing annotation effectively, it’s just as important to inspire people with specific and compelling use cases. These use cases help users understand that adding context can help elucidate meaning in a specific way for other readers.

Speaking of specific use cases, here are some other annotation tools we took note of during the conference: Wordnik and its API, Papero, Ponga, Hypothes.is. And we’d be remiss if we didn’t include a plug for Checkdesk, our software for verifying breaking news online.

Go check them out and let us know what you think… by leaving a footnote here.