Responding to the Forthcoming Demise of TwapperKeeper
Posted by Brian Kelly on 11 Dec 2011
Twapper Keeper Archive Service to be Shut Down
On 8th December 2011 the following announcement was made on the Twapper Keeper Web site:
Transition update
Twapper Keeper’s archiving is now available in HootSuite! As a result, we will be shutting down Twapper Keeper. Existing archives will be kept running until Jan 6, 2012, after which you will not be able to access your archives anymore.
Twapper Keeper has been widely used within the UK’s higher education sector, especially for archiving tweets containing event hashtags at events aimed at the developer, researcher and library sectors.
The popularity in the service has helped to demonstrate the importance of Twitter archiving, something which was not necessarily widely appreciated a few years ago. But in light of, for example, the recent news item on the JISC Web site which announced that “Social media ‘not to blame’ for inciting rioters” and went on to describe how:
A study of 2.4 million Twitter messages from the time of the riots has found that politicians and other commentators were wrong to claim the website played an important role in inciting and organising the disturbances.
we can see that the importance of Twitter archiving for a variety of purposes is now more widely understood. However it seems that Twapper Keeper will not be providing a long term repository of tweets. This does not necessarily mean that tweets will be lost since, as described in an article on Tweet Eternal: Pros and Cons of the Library of Congress Twitter Archive published in Time on 8 December 2011 “Thanks to a deal between Twitter and the United States Library of Congress, every public tweet sent on the social messaging service since its creation will become part of the Library of Congress’ digital archive, available to researchers and historians as an example of contemporary life and culture“. However as highlighted in Nature in n article on Social science: Open up online research “Social media hold[s] a treasure trove of information [but] the secretive methods of ethics review boards are hindering their analysis, says Alexander Halavais.”
Since it unclear when and if the Library of Congress archives will be made publicly available people and organisations which have made use of Twapper Keeper may wish to migrate the content of these archives. This post will describes approaches for migrating existing data, ways of identifying which archives may need to be preserved and ways of identifying key stakeholders who may need to make such decisions.
Migration of Existing Archives
Tools
Since creators and users of Twapper Keeper archives have less than a month to migrate their content, this post will outline ways in which the archives can be managed, and a discussion about the implications of the announcement of the closure of the service will be made at a later date.
Martin Hawksey has published a post on his MASHe blog which describes how you can Free the tweets! Export TwapperKeeper archives using Google Spreadsheet. Martin’s post also links to a post entitled LIBREAS.Library Grab your TwapperKeeper Archive before Shutdown! which describes a technique which can be used by those familiar with R code. Tont Hirst on the OUseful.info blog has also listed a technical solution based on R code in his post on Rescuing Twapperkeeper Archives Before They Vanish.
For people who may not be familiar with use of Google Spreadsheets or implementation of software applications for accessing Twitter archives you should note that you can also use a Web browser to view archives of interest (having ensured that all items are displayed and not just the default 10 items). You can then view the HTML source and save the file so that you have a HTML representation of the tweets which you can take manage locally. In addition, you can also save an RSS representation of the tweets which will provide a more structured format which should be more amenable to subsequent processing, if you wish to do this. Examples of this approach can be seen from the copies of the IWMW10 and IWMW11 archives.
Selection Criteria
In addition to being aware of the tools which can be used there will also be a need to decide which archives may be still be of relevance and identifying who may need to take responsibility for migrating the content to an appropriate location. Tony Hirst, in his post on Rescuing Twapperkeeper Archives Before They Vanish, has suggested that “one approach might be to look to see what archives have been viewed using @andypowe11′s Summarizr“. However although the Summarizr home page lists recently viewed Summarizr summaries of Twapper Keeper archives, it is not clear if a comprehensive list is available and, even if such a list could be made available, how this would inform decisions on the selection of archives to be migrated.
An alternative approach is to look at the TwapperKeeper archives which have been created by particular Twitter IDs. We can see, for example, that Tony Hirst (@psychemedia) has created 27 archives. Similarly using Twapper Keeeper’s search facility I find that I have created a total of 62 Twapper Keeper archives. Perhaps the initial stage in identifying archives to be migrated is for active Twapper Keeper users to identify the archives they have created, and then for them to make a decision of archives to be migrated, where the new archives are to be hosted and what to do for acrhives which will not be migrated, which might include informing key stakeholders.
Case Study
Rather than attempting to keep a copy of all of of the Twapper Keeper archives I have created, in this post I will provide a summary of the archives I created and docum the decisions I have taken regarding migration of the content and the reasons for these decisions.
Migrated to UKOLN Web site: The IWMW2009, IWMW10 and IWMW11 archives, which will be made publicly available, together the UKOLN and Ariadne_Mag archives which will be stored locally if we decide at a later date to analyse the tweets.
Key stakeholders informed: A number of archives may of interest to organisations such as JISC, CILIP, ALT, UCISA and CETIS. These organisations will be notified of the archives which I have created and will be informed of the techniques described in this post if they wish to migrate the content.
Archives of personal interest: Archives of personal tweets and personal interests have not been migrated.
Other archives: Other archives include archives for broad subject areas (e.g. #a11y, #phdchat) for which a general tweet about the forthcoming demise of the Twapper Keeper archive will be made and archives for events and areas of interest for which I had a short-term interest and wished to be able to view the tweets but which which I have no longer term interest.
A summary of the Twitter archives and the decisions I have made are given below. Please note that:
- The data given in the table was collected on 9 December 2011.
- The decisions given in the table may be changed at a later date.
- Twapper Keeper archives for other areas relevant to myself and UKOLN colleagues may have also been created. The #IWMW09 archive, for example, will be migrated and decisions about other archives will be made shortly.
Archive Type | Name | Description | # of Tweets | Create Date | Comment |
#Hashtag | #a11y | Accessibility (a11y) | 96,491 | 04-25-10 | #a11y community to be informed. |
#Hashtag | #a11yhack | DevCSI hack day | 329 | 06-21-11 | One-off DevCSI event. Report has been published. |
#Hashtag | #accbc | CETIS/BSI Accessibility SIG meeting | 396 | 02-28-11 | One-off DevCSI event. CETIS SIG coordinator to be notified. |
#Hashtag | #altc2009 | The ALTC 2009 conference | 4,754 | 08-28-09 | Large annual event. Report has been published. Event organisers to be notified. |
#Hashtag | #altc2012 | The ALT-C 2012 conference (Association for Learning Technology) | 104 | 09-12-11 | Created for next year’s event. Content not migrated. |
#Hashtag | #altmetrics | New approaches for developing metrics for scholarly research | 1,393 | 01-15-11 | #altmetrics community to be informed. |
#Hashtag | #Ariadne | The Ariadne hashtag – which may be used for UKOLN’s Ariadne ejournal. | 42,102 | 09-21-10 | Content not migrated due to multiple uses of tag. |
Keyword | Ariadne | Archive of tweets contains the string ‘Ariadne’ | 79,991 | 09-21-10 | Content not migrated due to multiple uses of keyword. |
@Person | ariadne_ukoln | Tweets about the Ariadne web magazine. | 2,792 | 05-28-10 | Content to be migrated to UKOLN. |
#Hashtag | #Bathcr | The University of Bath’s Connected Researcher activity. | 296 | 04-14-11 | #Bathcr community to be informed |
#Hashtag | #brdidc11 | Symposium on Data Attribution and Citation Practices and Standards, August 22-23 2011, Berkeley | 51 | 08-22-11 | Content not migrated. |
@Person | briankelly | Tweets about Brian Kelly | 9,952 | 03-19-10 | Content not migrated as alternative backup available. |
#Hashtag | #CETIS | The CETIS service, based at the University of Bolton. | 9,561 | 09-24-10 | CETIS colleagues to be informed. |
#Hashtag | #CILIP | CILIP, the Chartered Institute of Library and Information Professionals. | 14,356 | 09-24-10 | CILIP colleagues to be informed. |
#Hashtag | #CILIP1 | Campaign on future of CILIP organisation based on CILIP’s 1-minute messages. | 357 | 06-13-10 | Content not migrated. |
#Hashtag | #CSR | Comprehensive Spending Review | 0 | 10-15-10 | Content not migrated. |
#Hashtag | #dataprato | Invitational workshop to identify & agree areas for joined-up international action in research data management. | 128 | 04-11-11 | Content not migrated. |
#Hashtag | #digdeath | The conference on Death and Dying in a Digital Age held in Bath, UK | 72 | 06-25-11 | Content not migrated. |
#Hashtag | #eduwebconf | The eduwebconf conference | 33 | 11-07-11 | Content not migrated. |
#Hashtag | #falt09 | ALTC Fringe | 219 | 08-28-09 | Content not migrated. |
#Hashtag | #fbdevlove | The Facebook developers hack day | 1,297 | 03-26-11 | Content not migrated. |
#Hashtag | #fpw11 | The Future of the Past of the Web conference, British Library, London on 7 October 2011. | 755 | 09-22-11 | Event organisers to be notified. |
#Hashtag | #heweb10 | Tag for the HigherEdWeb 2010 conference | 8,812 | 09-28-10 | Content not migrated. |
#Hashtag | #heweb11 | The HighEdWeb 2011 conference, 23-26 October 2011 | 11,505 | 10-23-11 | Content not migrated. |
#Hashtag | #ILI2011 | Internet Librarian International 2011 conference held in London on 27-28 Oct 2011. | 3,067 | 10-27-11 | ILI organisers to be notified. Report has been published. |
#Hashtag | #ili2012 | Tweets for the Internet Librarian International (ILI) 2012 conference | 3 | 10-29-11 | Created for next year’s event. Content not migrated. |
#Hashtag | #ipres10 | Tweets for the iPres10 conference, Vienna, 19-24 Sept 2010. | 5 | 08-27-10 | Content not migrated. |
#Hashtag | #ipres2010 | Archive for the IPres 2010 conference to be held in Vienna on 19-25 Sept 2010. | 1,424 | 08-27-10 | Content not migrated. |
#Hashtag | #ISKB | A holder for the ISKB | 27 | 09-17-11 | Content not migrated. |
#Hashtag | #iwmw12 | UKOLN’s Institutional Web Management Workshop (IWMW) 2012 event | 2 | 10-29-11 | Created for next year’s event. Content not migrated. |
@Person | iwmwlive | IMWM live blogging account | 3,744 | 04-30-10 | Content to be migrated. |
#Hashtag | #jisc10 | JISC 2010 conference | 2,065 | 04-02-10 | Event organisers to be notified. |
#Hashtag | #jiscHTML5 | JISC HTML5 Case study work | 18 | 11-18-11 | Content not migrated. |
#Hashtag | #jiscpowr | Archive of tweets related to the JISC PoWR project provided by UKOLN and ULCC | 13 | 07-09-10 | Content not migrated. |
#Hashtag | #jiscpowrguide | Archive of tweets about the Guide to Web Preservation published by the JISC-funded PoWR project and launched on 12 July 2010. | 2 | 07-09-10 | Content not migrated. |
#Hashtag | #JISCPP | The JISC-Funded Patients Participate project. | 0 | 05-25-11 | Content not migrated. |
#Hashtag | #ldow2010 | Linked Data on the Web 2010 conference | 530 | 04-25-10 | Content not migrated. |
#Hashtag | #loveHE | Times Higher Education campaign to support Higher Education in UK. | 20,719 | 06-12-10 | Content not migrated. |
#Hashtag | #mdforum | UKOLN’s Metadata Forum | 1,746 | 12-10-10 | Content to be migrated. |
#Hashtag | #morris | Tweets about Morris dancing | 183,338 | 10-16-10 | Content not migrated. |
#Hashtag | #OAweek | Open Access week | 4,603 | 10-19-11 | Content not migrated. |
#Hashtag | #online11 | The Online Information 2011 conference held in London on 29 November -1 December | 3,915 | 11-29-11 | Content not migrated. |
#Hashtag | #oxsmc09 | socialmediaconference | 1,063 | 09-18-09 | Content not migrated. |
#Hashtag | #PhD | Tweets for researchers using the #PhD tag | 161,215 | 09-24-10 | Content not migrated. |
#Hashtag | #s113 | Workshop session at ALTC 2009. | 1417 | 09-03-09 | Content not migrated. |
#Hashtag | #scl2010 | Scholarly Communication Landscape (SCL): Opportunities and challenges symposium, held at Manchester Conference Centre on 30 November 2010. | 0 | 12-02-10 | Content not migrated. |
#Hashtag | #SHB11 | Security and Hunan Behavior conference | 1,117 | 06-18-11 | Content not migrated. |
#Hashtag | #SLG2011 | CILIP School Librarian Group conference. | 283 | 04-03-11 | Content not migrated. |
#Hashtag | #thatlondon | People (Northerners?) talking about going to “that London” | 1,781 | 07-09-11 | Content not migrated. |
#Hashtag | #ucassm | Social Media Marketing Conference organised by UCAS. | 225 | 10-18-10 | Content not migrated. |
#Hashtag | #ucsoc12 | UCISA SSG (Support Services Group) event. | 5 | 09-05-11 | Content not migrated. |
#Hashtag | #udgamp10 | What Can We Learn From Amplifed Events seminar, given by Brian Kelly, UKOLN at the University of Girona | 395 | 09-01-10 | Content migrated. |
#Hashtag | #ukmw09 | UKMuseumsandtheWeb | 750 | 12-05-09 | Content not migrated. |
Keyword | ukoln | Tweets about UKOLN | 3,385 | 03-19-10 | Content to be migrated. |
#Hashtag | #ukolneim | UKOLN’s Evidence, Impact, Metric work | 523 | 11-05-10 | Content to be migrated. |
#Hashtag | #UKOLNseminar | UKOLN seminars | 69 | 04-01-11 | Content to be migrated. |
#Hashtag | #UniofBath | Tweets about the University of Bath | 1,798 | 06-15-11 | Content not migrated. |
#Hashtag | #UniWeek | The UK’s Universities Week campaign. | 1,767 | 06-15-11 | Content not migrated. |
#Hashtag | #Virtualfutures | The Virtual Futures conference | 2,216 | 06-18-11 | Content not migrated. |
#Hashtag | #w3ctrack | W3C Track at WWW 2010 conference | 205 | 04-30-10 | Content not migrated. |
#Hashtag | #W3CUKI | W3C UK and Ireland Office | 266 | 04-18-11 | Content not migrated. |
#Hashtag | #ww2010 | Misspelling of WWW2010 hashtag | 904 | 04-29-10 | Content not migrated. |
I welcome suggestions on other tools and approaches which can be used for managing such archives and also approaches to selection and deletion criteria for Twitter archives.
Chris Rusbridge said
An wakward thing is that twapperkeeper archives are often not created by the obvious person. I think the #idcc11 tag was created by someone in California, for example. And it’s also a problem that tags can’t necessarily be recreated post hoc, as it’s a bit serendipitous which tag gets taken up. Was it #idcc09 or #idcc5, for example? If I had kept my own tweets it would be easier to tell.
Brian Kelly (UK Web Focus) said
Yes, that’s an interesting consequence of the distributed nature of both minting Twitter hashtags and created archives of such hashtags. I suspect the management of the archives will be more difficult if the community (or activists within the community) are ahead of key stakeholders, such as the event organisers.
For info, here are links and summaries for the IDCC conferences from 2009-2011. Note that I am unfamiliar with the Twitter IDs of the people who created the archives:
Twitter Archiving Revisited: Preparing for the Demise of Twapperkeeper :: Jennifer M Jones said
[…] decide to allow access to those many years of tweets they are accumulating records on. As Brian Kelly notes it is unclear when that might be and how much access will be allowed to the average researcher […]
Free the tweets! Export TwapperKeeper archives using Google Spreadsheet – MASHe said
[…] share so that we can free the tweet!Update: Having read Brian Kelly’s post on the Responding to the Forthcoming Demise of TwapperKeeper it occurred to me that some coordination might be required. So if you use this template could you […]
Tony Hirst said
Brian – I started working on a Python script that will download all the archives created by a particular individual, including large Twapperkeeper archives. The code is at: https://gist.github.com/1463711
Rescuing Twapperkeeper Archives Before They Vanish, Redux « OUseful.Info, the blog… said
[…] a post fron @briankelly (Responding to the Forthcoming Demise of TwapperKeeper), where Brian described how to lookup all the archives saved by a person on Twapperkeeper and using […]
TwaperKeeper archives | nostuff.org said
[…] Anyway, Martin Hawksey has created a wonderful tool for archiving, ummm, archives before they are gone. The tool is actually a Google Spreadsheet and to me it’s a testament to Google Docs power that an application that fetches and stores data from another site can be created using it. Brian Kelly has more info here. […]
Chris Keene said
Archives I’d created on TwapperKeeper can now be accessed from here:
http://www.nostuff.org/words/2011/twaperkeeper-archives/
Chris
Steve Hitchcock said
Brian, I notice neither you or any of the possible archiving solutions you have reported mention HootSuite, the service linked in the first sentence of the TwapperKeeper notice. Do you have any thoughts on that? Obviously many people will follow that link. I did and it wasn’t clear how it replaced TW’s archiving service.
Brian Kelly (UK Web Focus) said
Hi Steve
In a post entitled No More Access To Your Twapper Keeper Archives Gary Green has raised a number of questions about Twapper Keeper integration with HootSuite. he also raised these questions on the HootSuite blog but didn’t get a satisfactory response:
Hi Gary,
Thanks for stopping by. Our developers are currently exploring options integrating TwapperKeeper. Let us know what you’d like by visiting our Feedback Channel.
Simon Dobson, JISC programme Manager then asked:
Please could you provide me with a clear answer to the following question: will current TwapperKeeper archives be preserved and made available via HootSuite or will current users who want to retain their archives need to extract them before the closure of TwapperKeeper?
but also failed to receive a satisfactory answer, with the announcement of Twapper Keeper’s demise simply being repeated:
Twapper Keeper’s archiving is now available in HootSuite! As a result, we will be shutting down Twapper Keeper. Existing archives will be kept running until Jan 6, 2012, after which you will not be able to access your archives anymore.
Thanks for using TwapperKeeper – we look forward to seeing you at HootSuite.
From one perspective it would appear that HootSuite have no interest in the higher education sector, or the research interests we have in making use of Twitter archives. Perhaps they will be aimed at the consumer market?
On the other hand, there may simply be problems with internal communications which are the reasons for the lack of information coming form HootSuite. I will post an update on this blog if I hear of any further developments.
Twapper Keeper: going, going….. (almost gone) | DMU MashedLibrary said
[…] What are the alternatives? Martin Hawskey discusses how to Export TwapperKeeper archives using Google Spreadsheet, while Brian Kelly has further advice on selecting and migrating archives. […]
Chris Keene said
Any suggestions for what to use in future instead of twapperkeeper?
Goodbye Twapper Keeper « The Event Amplifier said
[…] Kelly’s recent post, Responding to the Forthcoming Demise of Twapper Keeper, outlines the processes currently available for migrating Twapper Keeper archives and suggests some […]
Alternatives To Twapper Keeper « UK Web Focus said
[…] On 23 December I received an email which confirmed the news about the forthcoming demise of the Twapper Keeper Twitter archiving service: […]
The #LODLAM Session at #SXSW Demonstrates Importance of Consistency of Session Hashtags « UK Web Focus said
[…] were invented on 23 August 2007. Their role in events quickly became apparent and by 2009, as referenced in a post about Twitter archiving, event hashtags were being used at large events such as #ALTC2009 and #IWMW2009. In August 2009 […]
Collecting Echoes | Digital Information & Technology Adventures said
[…] There are plenty of online twitter archiving and analytics platform that hide the magic behind software as a service (SaaS), such as Tweet Archivist, those these can require payment and there remains the question mark over what happens when the service is withdrawn, like with the demise of Twapper Keeper. […]