Learning From Shared Twitter Links (Before Trunk.ly’s Demise)
Posted by Brian Kelly on 9 January 2012
The Forthcoming Demise of Trunk.ly
On 19th February 2011 I signed up for the trunk.ly service. The email I received which confirmed my registration summarised the features of this service:
- Trunk.ly indexes the full web page that all your links point to. Just search and find, no need to worry about tagging or summarizing content.
- If you #tag content in Twitter, or tag it in Delicious, Trunk.ly will create tags for you.
- Trunk.ly also checks your Twitter favorites so you can just favorite content with links without retweeting it if you prefer.
I have to admit that I’d forgotten about trunk.ly until I received an email recently telling me that the service has been acquired by AVOS (who have recently acquired Delicious.com) and that the trunk.ly service will terminate from the end of the week: Friday 13th January.
The email did inform me that I can export the content created by trunk.ly:
This tool creates a list of all your bookmarks in a format understandable by most browsers. You can save the generated page (as HTML) and import it into your browser — or anything else that accepts bookmarks in a standard format.
Your tags will be included in the export file even if you don’t see them on the page. This is the limitation of the export file format.
I have exported the content and hosted it on the UKOLN Web site. However before the service is withdrawn I thought it would be useful to explore what it can tell me about the links I have shared on Twitter.
The service is associated with my main Twitter account (@briankelly) and with the UK Web Focus blog. Since registering with the service ten months ago it has harvested 4,997 links. I am followed by six other trunk.ly users and follow 13 users.
The service allows me to browse through the links I have created in chronological order as well as the links created by people I follow. As illustrated Trunk.ly can summarise the content of the link and, if available, include an embedded image.
Trunk.ly also allows me to explore the content by any associated tags.
As shown in the accompanying screen image I can see that I used the #altc2011 Twitter hashtag for a number of tweets. Clicking on the tag enable me to view the three tweets I posted: one which linked to a FriendFeed post in which Seb Schmoller described how ““Recording can improve a bad lecture! 7… – Seb Schmoller – FriendFeed”; one on ““Battling legal, logistical and technical obstacles to archiving the Web” « UK Web Focus” which summarised one of my blog posts on Twitter archiving and one on “Martin Hamilton’s blog: ALT-C 2011: Cloud Learning with Google Apps” in which I retweeeted Martin Hamilton’s link about a presentation he gave at the ALT-C 2011 conference.
Of more interest, however, is Trunk.ly’s search interface. This enable me to search not only resources which I have shared but also resources shared by the people I follow as well as all Trunk.ly users. Examples of the terms contained in links posted by myself and Tony Hirst (@psychemedia) are given below:
|Search term||Domain search|
Unsurprisingly we both tweet significant numbers of links back to our host institutional Web site.
It is also possible to search by the resource type which have been shared:
In February 2009 Mike Ellis that, for services such as Twitter and blogs “The person is the point“:
Twitter, like blogging, needs an edge, a voice, a riskiness. As long as institutions can retain this – i.e., do it for a reason – then, IMO, things will get more interesting. If they don’t, we’ll probably all be unfollowing museums as quickly as we can slide down the steep, slippery trough of disillusionment…
That may have been the case in Twitter’s early days but now Twitter does not need to have an edge. Twitter can be used for sharing ideas and resources and for discussing the implications of the ideas and commenting on the resources.
The Trunk.ly blog has announced that:
Trunk.ly will be discontinued, and we will immediately start working to integrate our technology and insights to accelerate the link-saving and searching capabilities in Delicious.
I’m pleased that I still have my Delicious account and will be interested to see how the service becomes embedded within Delicious. It will also be interesting to see if the resource sharing capabilities provided by Twitter, and the ways in which such sharing can now be analysed will have a role to play in the development of altmetrics. As described in the altmetrics manifesto:
Articles are increasingly joined by:
- The sharing of “raw science” like datasets, code, and experimental designs
- Semantic publishing or “nanopublication,” where the citeable unit is an argument or passage rather than entire article.
- Widespread self-publishing via blogging, microblogging, and comments or annotations on existing work.
The tweet provides a link to a paper on “Can Tweets Predict Citations? Metrics of Social Impact Based on Twitter and Correlation with Traditional Metrics of Scientific Impact” which concludes:
Tweets can predict highly cited articles within the first 3 days of article publication. Social media activity either increases citations or reflects the underlying qualities of the article that also predict citations, but the true use of these metrics is to measure the distinct concept of social impact. Social impact measures based on tweets are proposed to complement traditional citation metrics. The proposed twimpact factor may be a useful and timely metric to measure uptake of research findings and to filter research findings resonating with the public in real time.
These conclusions were based on analysis of all tweets containing links to articles in the Journal of Medical Internet Research (JMIR). For a subset of 1,573 tweets about 55 articles published between issues March 2009 and February 2010, different metrics of social media impact were calculated and compared against subsequent citation data from Scopus and Google Scholar 17 to 29 months later. A heuristic to predict the top-cited articles in each issue through tweet metrics was validated.
For those working in the area of medical internet research it would seem that Twitter has an important role to play in increasing citations or helping to identify important papers. Perhaps, after all, Mike Ellis is right: the person is the point. But the person may be the researcher and the point may be the research, rather than the researcher’s edgy voice.
Survey Paradata: As described in a post on Paradata for Online Surveys blog posts which contain live links to data will include a summary of the survey environment in order to help ensure that survey findings are reproducible, with information on potentially misleading information being highlighted. The survey findings described in this post were collected on 30 December 2011 using the Google Chrome browser on a PC running Windows 7. It was noticed that there were differences between the two ways of finding the numbers of links which have been harvested: the information provided in the user’s profile (e.g. see my profile page which states that there are 4,997 links) and the numbers given for a search for the user (see my search results).