UK Web Focus (Brian Kelly)

Innovation and best practices for the Web

A Tale of Three Conferences

Posted by Brian Kelly on 7 Dec 2009

Three Related Conference in One City

Last week, as I mentioned previously, I attended the Online Information 2009 conference, held in Olympia London on 1-3 December. But this wasn’t the only conference of interest to me which took place in London last week – the 5th International Digital Curation Conference was organised by colleagues of mine at UKOLN and the UK Museums and the Web conference is an event I have spoken at in previous years.

But as well as the content of these conferences being of interest to me, these conference also made extensive use of Twitter, which enabled engagement with the conference discussions to include people who weren’t physically present and allowed the conference outputs and discussions to be read and analysed afterwards. In this post I provide a summary based on statistics of the use of Twitter at these events and suggest that we will need to explore ways in which misuse of event hashtags (by Twitter spammers) can be tackled.

The Online Information Conference

As might be expected for an international conference aimed at information professionals the event had a conference hashtag (#online009) and, despite problems with the WiFi network, according to Twapperkeeper archive for the #online09 tag, there were 2,351 tweets published between 22 November and 5 December 2009.

Additional statistics are provided by the WTHashtag service. The accompanying image was created by this service.

I should also add that I used the Tweetwally service to create a ‘Tweetwall’ of the event’s tweets – but it seems that this only displays tweets posted in the past few days.

The IDCC Conference

The 5th International Digital Curation Conference (organised by DCC – the Digital Curation centre – which included colleagues of mine at UKOLN) was held on 3-4 December 2009. As described on the Digital Curation blog this was an amplified event, with an event hashtag (#idcc09), a live blogger (the @idcclive Twitter account) and a live video stream. According to the Twapperkeeper archive for the #idcc09 tag, there were 782 tweets published between 2 and 5 December 2009.

Additional statistics are provided by the WTHashtag service. The accompanying image was created by this service.

The UK Museums and the Web conference

Finally the one-day UK Museums and the Web conference was held on 3 December. On this occasion According to the Twapperkeeper archive for the #ukmw09 tag 706 tweets were posted between 27 November and 5 December 2009.

Additional statistics are provided by the WTHashtag service and the accompanying image was created by this service.

Exploring The Twitter Communities

Detecting Spammers

We have some statistics which seems to indicate that Twitter has played a significant role in supporting these three events.But might these raw statistics be skewed by Twitter misuse, such as Twitter posts from spam followers? In order to seeks an answer to this question I have made use of Tony Hirst’s software to analyse Twitter communities centred around an event hashtag.

As the software is based around a combination of a Twitter user as well as the event hashtag I had to chose a Twitter user likely to have a wide following in order to explore the community tweets. I used myself (@briankelly) for the #online09 conference and Mike Ellis (@m1ke_ellis) for the #ukmw09 conference. I had intended to use Chris Rusbridge (@cardcc) for the #idcc09 conference but no results were provided for this twitter ID so I used the official event live blogger account (@idcclive) instead.

You can view the findings for the #online09 conference; #idcc09 conference and #ukmw09 conference.

Tony’s software does correctly identify spammers on the event hashtags. Some, such as @ProvidenciaAmar have already had their account suspended whilst others,such as the helpfully labelled @MommyIsSoSexy ID is still available, and can be easily identified as a spam account. But have other Twitter accounts been incorrectly labelled as spam accounts, I wonder?

Conclusions

Some of the early adopters of Twitter felt that Twitter was very much about the individual and was ;of the moment’, with no need for archiving tweets for reuse or analysis. I think this is no longer true – or, rather, this is no longer the only use case for Twitter. In the case of use of Twitter to support events we are definitely seeing people wishing to view the tweets afterwards. In addition the popularity of Twitter at events has its downside – and we are seeing an increase of Twitter spam, with inappropriate content and links being labelled with popular event hashtags.

Tony Hirst’s software has made some initial steps in exploring ways of automatically identifying Twitter spammers. I suspect that such techniques will soon be embedded in Twitter tool. But since Tony’s approach is based on Twitter users which have been connected with a trusted user, this approach will not necessarily work for events with a more distributed network, with no well-established Twitter ‘hubs’. I wonder if an official event Twitter account might provide such a hub, allowing users to follow the account in advance of the conference. At the IWMW 2009 event we used two official Twitter accounts: iwmw and iwmwlive and the live-blogging Twitter account was also used at the IDCC conference (indeed the idcclive Twitter account was managed by the same person, Kirsty McGill).

What do you think: a sensible development or unwanted complexity liable to stifle Twitter’s flexibility and informality?


Twitter conversation from Topsy: [View]

5 Responses to “A Tale of Three Conferences”

  1. Les Carr said

    What’s surprising about these results is the relative similarity of the sizes of all three Twitter streams. Online 2009 is a HUGE exhibition, so it should dwarf the contributions to others, if Twittering events was a mainstream phenomenon. As it is, it does seem to have a much longer tail, and less per capita twittering.

    Compare this to @eddieizzard, who has the audiences twittering before the start of every gig in his current tour.

    • Yes the Online Information 2009 conference was larger than the other two events. However there are other factors to consider besides the numbers of participants including:

      o The WiFi network was very poor at Online, which constrained the numbers of tweets.
      o The Online Information conference attracted significant numbers of participants from overseas. Twitter usage may be lower in mainland Europe than in the UK and the US. In addition people from overseas may have been reluctant to use their phone to tweet, due to the costs.
      o The Online Information conference attracted significant numbers from the commercial sector. These may be less inclined to tweet at events (apart from those who work in marketing!)

  2. We discovered another issue at IDCC 09, in that approximately 30 #idcc09 tagged tweets from the @idcclive did not appear in the Twitter search, and are therefore not included in your statistics above. When we looked into this during an interval and found a list of reasons here as to why tagged tweets might not be appearing in Twitter search. The only explanation we could deduce from this is that my tweets had been “filtered out of our search index for quality reasons” because of possible topic abuse – although the account itself was not suspended. As a dedicated event account, every tweet up until this point had included the #idcc09 tag. On the off-chance this was causing the problem, I tweeted once without the hash tag (athough included the term “IDCC 09” in the tweet) and within half an hour or so my tagged tweets were appearing in the search again.

    Realising that the @idcclive #idcc09-tagged tweets were not appearing in the search did change my behavior for the duration, as I we were having network problems as well, so I focussed instead on writing up blog posts about the sessions involved to be posted later at the event blog, rather than tweeting in so much detail. Whilst there were people following @idcclive, part of the way we were using the account – particularly as part of our netvibes page, where the live video stream appeared – really necessitated the updates appearing in the #idcc09 search.

    This is an issue I need to look into further – but wondered if you had any thoughts?

  3. I’m afraid I don’t have the stats to the detail that you’ve provided here, but the recent Middlemash event (http://middlemash.wordpress.com) struck me as more ‘densely’ tweeted than most (i.e. a higher percentage of overall attendees were contributing to the twitter stream). Because of the format of the event (half talks, and half practical doing stuff) I think the use of twitter was slightly different – for example I used it to share links with a group of people I was working with. This blog post from a non-attendee adds anecdotal evidence of the usefulness of the twitter activity http://kwiddows.blogspot.com/2009/12/did-anyone-else-find-benefit-from.html

    The question of archiving tweets bothers me not because I think this is not necessary – some tweets may be worth archiving – but because I’m not convinced that blanket archiving of tweets is useful. If we are serious about ‘archiving’ tweets for the longterm they need curating – not just capturing. (I tend to think this is the biggest challenge for digital preservation generally – deciding what to preserve)

    We did archive the tweets with Twapperkeeper (http://twapperkeeper.com/middlemash) and we’ve already had some people building on this – e.g. http://spreadsheets.google.com/pub?key=tV6ZiW6ZnhTUd-Gp-wg1z4g&single=true&gid=0&output=html – I think this is possibly the start of ‘curation’ although at a basic level (not all links will be of equal value/interest).

    Obviously any activity like this takes a significant amount of effort.

  4. […] https://ukwebfocus.wordpress.com/2009/12/07/a-tale-of-three-conferences/ In this blog posting Brian provides a comparative analysis of the Twitter activity of Online 2009, taking into account activity at two other conference that took place in the same week. […]

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

 
%d bloggers like this: