TwitterWeave: new IDS web app tracks research messages

22 March 2016

TwitterWeave - - is our new web application (app) for tracking exchanges about research on Twitter in real-time. Launched today and free-to-use, TwitterWeave is designed for researchers, communicators, campaigners and anyone who wants to learn more about how people engage with research. It is a tool for generating narratives about what is being discussed, when and by whom.

Holding a phone showing the Twitter icon

What has Twitter got to do with research engagement?

Social media platforms have long been used to raise awareness of, and discuss, all human activity – including research. And there is growing appreciation of the power that platforms such as Twitter and Facebook have to influence large numbers of people around the world.

The ability to accurately mine the data within the conversations that take place on Twitter offers a unique opportunity for researchers, information professionals and communications specialists to track the propagation of research messages and ideas across social media and gain insights into how particular people, professions and organisations are engaging with particular kinds of research.

There is increasing pressure on researchers from donors and the wider public to demonstrate the impact of research. But as we know, impact is often extremely hard to prove. TwitterWeave does not claim to solve this problem, but by revealing how people and institutions engage with research it can support efforts to show impact. It also offers a type of alternative metric because unlike most other approaches to metrics it doesn't focus on downloads or citations, but on the people who are engaging with the research.

A public good

While, much of Twitter's online chatter is publically available, social media data can be highly challenging and time-consuming to collate and interpret. Not only does the sheer quantity of data represent a major hurdle (it is estimated that 500 million tweets are sent every day) but identifying and interpreting patterns within that data is a complex process.

Powerful social media monitoring services which try to address these issues already exist, but if they're any good they tend to be expensive. A licence for such systems can easily cost around £10,000 per year and even if you can afford the price, the sheer power and complexity of these systems means that mastering them can be a slow process. By comparison, TwitterWeave is simple to use and comes with no fees at all.

Figure 1: TwitterWeave visualization of tweets

TwitterWeave visualisation of tweets

Putting people at the heart of the analysis

TwitterWeave maps the propagation of tweets over time and differently from many other Twitter apps it maps not only what is being said, but who is saying it. By displaying connected threads of tweets, retweets and replies, TwitterWeave shows how messages are relayed and responded to as they travel across the social space.

For example, tweets about a particular piece of research may originate from Twitter users based in academia, but within a few minutes or hours those tweets may then be taken up and retweeted within activist, practitioner or policy circles. TwitterWeave enables this tracking because it takes advantage of the way that Twitter users (both individuals and institutions) often include brief details about their roles and interests in their profiles. The app highlights the points in the thread when ideas pass from one community of users to another.

The number of followers for each Twitter user is another useful piece of information. A tweet from a personal or institutional Twitter account with 100,000 followers evidently has a much bigger potential audience than a tweet from an account with only 100 followers. Not only that, but the person or institution with 100,000 followers is likely to have significant influence beyond Twitter, because usually only individuals or organizations which command a great deal of respect and attention in the off-line world have large numbers of Twitter followers.

So the number of followers is used by TwitterWeave as a proxy for influence.

Figure 2: TwitterWeave thread of retweets and replies

TwitterWeave thread of retweets and replies

How do you use TwitterWeave?

Simply sign-in using your own Twitter credentials and then search for whatever you want, just as you would with Google. Based on your chosen keywords, TwitterWeave generates a graph representing all the tweets which match your search. The tweets are plotted as dots on a timeline and they are also ranked according to follower numbers (the higher a tweet appears on the graph, the more followers the author of that tweet has).

Clicking on a tweet opens a second graph which shows the Twitter ‘thread’ within which that tweet lies, i.e. the series of retweets and replies which contains that particular tweet. Selecting any of these points will display the content of the tweet, together with all available information about the author from his or her profile. All the information from each search is also reproduced in a downloadable spreadsheet file (JSON and CSV), for those wishing to do further analysis off-line.

We’d love to  hear your reactions to TwitterWeave so please let us know what you think!

Image credit: Esther Vargas (cc on Flickr)