Streamosphere: Interview with Euan Adie

Euan Adie in June announced the Streamosphere service on the Nascent blog: Welcome to the Streamosphere. His simple explanation of what Streamosphere does:

Streamosphere lets you track scientific discussion on the web, in real time.

The Nascent blog post explains Streamosphere in more detail, in another post from July Euan talks about some updates to the service. I've asked Euan a few questions to learn more about Streamosphere.

1. Can you explain what Streamosphere is and does?

Streamosphere is an aggregator of scientific activity on the web. It tracks what scientists are talking about on Twitter, wikis, blogs, social bookmarking services, forums and other web 2.0 services.

2. Why do we need such a tool?

I think scientific attention (any time you read an abstract, cite a paper or search for a particular gene in a database you're giving it attention) is valuable to aggregate. Traditionally scientists and publishers have measured attention with citation counts and to a lesser extent downloads but there's lots of other online activity – like blogging, commenting and bookmarking – that could be used in new metrics.

At the moment the relevant activity is

• distributed people could be talking about a paper I'm interested in on Friendfeed, blogs, forums…
• noisy there are tens of thousands of research scientists on Twitter but most of the time they're posting pictures of lolcats
• lacking provenance it's not easy to know whose opinions to trust

Streamosphere attempts to tackle these issues by aggregating activity from lots of different sites, removing spam, providing filters and disambiguating users on different services. If I'm interested in what other scientists think about a particular paper then I can search for it on Streamosphere. If I want to keep track of everything related to a particular field then I can do that too.

3. What is the difference between Streamosphere, Postgenomic and Nature.com Blogs?

Postgenomic and Nature.com Blogs only aggregate and analyze blog posts. Streamosphere gets a feed from Nature.com Blogs but covers other services too.

4. What is the difference to other aggregators such as FriendFeed?

The main difference is that you can't interact with content on Streamosphere – if you want to comment on a blog post or paper you have to visit it and leave your comment there.

Another would be that you don't follow people on Streamosphere, you'd follow topics.

5. How do you decide whether something or someone is related to science?

If you bookmark or discuss things with DOIs, arxiv IDs or PubMed IDs then you're added to the database. At this point any of your activity that doesn't look scholarly doesn't get aggregated.

After that it's down to manual checks and whether or not you're following / being followed by other people in Streamosphere.

6. Can we follow Streamosphere via RSS or Twitter?

Not yet! I'm hoping to open up the data in Streamosphere via a public API that can output RSS and other formats but don't want to do it before the system is stable. There are also questions surrounding how some messages can be republished. I'm not sure if Streamosphere's API output can include the contents of any tweets, for example, as to receive streaming updates from Twitter you need to sign an agreement saying that you won't distribute them outside of your application (can't remember the exact wording).

The best way to track updates about Streamosphere is to watch Nascent.

7. What are your responsibilities at Nature.com?

I'm a product manager in the web publishing group. Generally speaking the areas I'm responsible for are aggregation, the Nature.com Blogs homepage and mobile platforms.

8. What did you do before working for Nature.com?

I was a bioinformatician at the University of Edinburgh in the medical genetics unit, working on candidate disease gene prioritization.

Before that I'd co-founded and worked on a start-up that tried to extract entities and sentiment from forum posts. We brought in very little revenue for eighteen months, ran out of funding and went bust in 2001. It was an education…. 😉

9. Do you want to talk about future plans for Streamosphere?

Sure – the main thing I'm looking forward to is personalization – being able to log in and tell the system what you're interested in. From then on whenever you visit the site particularly relevant items will be highlighted.

The API is another key development… I'm really interested in seeing what other people can do with the data.