Chris Keene

Jan 12

"Social media ‘not to blame’ for inciting rioters"

Just before christmas my email and twitter feed was full of an announcement by the JISC titled: "Social media ‘not to blame’ for inciting rioters". The JISC basically fund technology and information projects/infrastructure in UK Higher Education. A lot of what they do impacts on my job so I follow various email lists and twitter accounts around their work.

But this heading struck me a little, it seemed a bold claim to make.

You can see the announcement here:

http://www.jisc.ac.uk/news/stories/2011/12/riot.aspx

A few things niggled me in the announcement: It only referred to twitter; it only linked to a Guardian article and it quickly moved on to assert that Twitter was in fact very helpful in organising the post riot clean ups. The latter is obvious, and easy to prove, but has nothing to do with the original assertion, that Social Media did not play a key role in the riots.

The "find out more" link at the bottom of that announcement takes you to this page:

http://www.jisc.ac.uk/whatwedo/programmes/inf11/einf/neiss.aspx

This is a standard JISC project profile page. The JISC do many wonderful things but their project pages are often of little use. Often I find myself wanting to find the report/sofware/conclusions/information that a project has produced, Google the project, find the project page, and all you see is some blurb in the future tense and a project plan. Not what I want!

In this case, you have to look you have to look half way down the right hand side (not at the bottom where I expect to find more info) to find the homepage of the project. And that takes you to:

http://drupals.humanities.manchester.ac.uk/neiss3/

The only reference to the study in question is at the top right, a headline to a very brief news item "twitter did not incite rioting", which links to a Guardian article (but not the main one).

The main Guardian article is here:

http://www.guardian.co.uk/uk/2011/dec/07/twitter-riots-how-news-spread?newsfeed=true

I have a few issues:

The Guardian article starts with a link 'get the data' - I know it sounds petty but it's wrong. The article talks about millions of tweets, the data you can get is just a list of popular usernames and hashtags during the riots, very different.
"Relevant tweets were drawn from dozens of riot-related hashtags – such as #EnglandRiots or #BirminghamRiots – which were used at the time to pool tweets about the same subject" - I'm not an expert, but surely many of the people involved would not have been using a hashtag. Armchair commentators and wannabe wits (HELLO!) do use them. If this is true, then studying tweets from the latter and concluding that they weren't inciting riots is flawed.
Likewise, no mention is given as to if this dataset included private (locked) twitter accounts, or if it included DMs, an important point, as those with more than two brain cells might be more careful what they tweet on a public feed as opposed to DMs or a private account.
You could argue that these are technicalities, not suited to a newspaper article. Fine, I would normally check the source. But I couldn't find a link anywhere to a full report on the study. It seems that the Guardian article was the write up.
This was a study in to whether Twitter incited the riots, not Social Media in general, as some of the articles claimed. In fact from what I remember, it was Blackberry and Facebook which were mentioned the most during news reports at the time (both get a passing reference in the article). You can not conclude anything about social media's role in the riots just by looking at Twitter, the headlines like that I quoted in the title of this post are misleading.
As noted, the study quickly moves on from the point in question to pointing out that twitter helped initiate the cleanup operation. This starts to feel like an agenda (we must show how wonderful social media is and how wrong the daily mail are), we should be better than that.

I'm not trying to have a go at the researchers, JISC or The Guardian. But this seems like a bold statement to make which doesn't stand up to scrutiny (though it's hard to say for sure with little to verify). Any claim about social media's role needs to look at blackberry messenger messages and facebook communication, even if these may be hard to obtain. To be blunt, I'm sure kids communicating with each other - via social media - about where to go and encouraging mates to come along probably did contribute.

I love social media, and certainly don't think it should be 'turned off' when trouble occurs but this leaves me wanting.

Sep 13

Electoral Boundaries for Brighton

I've taken a quick look at the proposed boundary changes for Brighton and Hove. Currently there are three constituencies, formed of three column like blocks each running from coast to northern city boundary in a row. They are: 'Brighton Pavilion', the central part of Brighton, Brighton Kemp Town, the eastern side, and Hove (actually).

The new constituencies look like this:

- Brighton and Pavilion and Hove

http://rr-bce-static.s3.amazonaws.com/wp-content/uploads/2011/09/Brighton-Pavilion-and-Hove-BC.pdf?9d7bd4

- Brighton and Hove North

http://rr-bce-static.s3.amazonaws.com/wp-content/uploads/2011/09/Brighton-and-Hove-North-BC.pdf?9d7bd4

Lewes is currently with the coastal towns of Seaford and Newhaven. The new proposals split this and put Lewes with Eastern Brighton

http://rr-bce-static.s3.amazonaws.com/wp-content/uploads/2011/09/Lewes-and-Brighton-East-CC.pdf?9d7bd4

(Seaford and Newhaven are in a strange new ward which runs from the coast all the way up to near the border with Kent)

http://rr-bce-static.s3.amazonaws.com/wp-content/uploads/2011/09/Uckfield-CC.pdf?9d7bd4

These proposals link liberal (small l), radical, green (we have the first and only Green MP!) central Brighton with conservative Hove. Not good bed fellows, though it is mainly the more hip coastal/central parts rather than the suburbs.

B&H North has some logic, putting the suburbs together in one large bundle (who tend to have a different politics to the younger, more transient centre), HOWEVER, also put in to this constituency is the University of Sussex campus. I'm not sure our young lefty radicals will take to that!

The Lewes constituency has been a dot of yellow in a sea of blue. It becomes a very odd constituency, part of the city of Brighton (normally either labour or tory), Lewes (lib dem), and a fairly good chunk of rural East Sussex (very tory) as well. Norman Baker may well have his work cut out. In any case this constituency in particular seems a compromise, mixing rural, city and town in to one.

You can see all the proposed constituencies for East Sussex and Kent here:

http://consultation.boundarycommissionforengland.independent.gov.uk/whats-proposed/south-east/brighton-and-hove-east-sussex-kent-and-medway/

Aug 23

openurl.ac.uk data

I started to play with the openurl.ac.uk dataset http://openurl.ac.uk/doc/data/data.html - inspired by the Discovery competition run by "UK Discovery and the Developer Community Supporting Innovation (DevCSI)". Unfortunately, partly due to this being summer, I didn't get much time to work on it, and only the most bare skeleton is up and working:

http://nostuff.org/openurlde/openurl1/

This tried to provide some general usage stats based on the data. Next steps would be to include time based data, e.g. when is a particular journal/article/source popular? Secondly, a compare two or more journals/articles/etc. Thirdly graphs and charts for everything! fourthly, make it all pretty and dashboard-like. There's also potential to bring in data from other sources, and link out, especially for journal titles.

More cynically, it would be interesting to try and reverse engineer the institution resolver id to University name.

The code is at https://github.com/chriskeene/openurl1
And it's quite simple to try it yourself.

I used netbeans as a IDE, git and github for code tracking/sharing. This is the first time using all three of these, so there was a bit of a learning curve as well as just diving in to write code.

I also made a bit of a start with another idea, to provide a service for searching publishers, especially those smaller publishers who were less well known. It would then show books published by that publishers. My timeline for working on it went a little like this: see email in early July about some sort of competition. A week later actually get around to reading it properly and realise competition ends when July does, put socks on, start to work on it trying to use Cambridge and BL/Bnb datasets. SPARQL endpoint at Cambridge timed out when doing my (simple) search for publishers. BNB had no sparql endpoint according to ckan (later the record was updated to show an endpoint), and the dataset was only a small portion of bnb. Realise only have a week left to end of July and no time to make any more progress so give up (but do find a endpoint for bnb, required very different sparql query to get required data, but didn't timeout which was nice). A week later the competition re-opens, but after a little more playing decide to move on to something else. you can see the code, as it is,here https://github.com/chriskeene/discovery1

Aug 1

Southern Brighton Mainline, single point of failure

Tonight there is massive disruption on the Brighton Mainline. The line serves the whole of East and West Sussex, far more than the name suggests. And it's incredibly busy, the sort of line where trains are described as being like like cattle trucks - especially in peak hours.

East Croydon is the lynch pin of this line. While Clapham Junction may be more (in)famous, it caters for South West Trains from Waterloo and most of Souther running from Victoria, trains from London Bridge go no where near it, so a major problem could occur and trains will still get from London to the coast.

All trains must travel through East Croydon. So when a major mudslide due to a burst water main occurs just south of East Croydon it's big.

[image copyright ATOC / nationalrail.co.uk]

So what happened? They did what they normally do at these times, laid on rail replacement buses.

As you can imagine, the local bus companies don't tend to have lots of spare staff and buses sitting around the depot in peak hours, so this is alway a slow and limited response. Queues at Easy Croydon ran (apparently) for miles and waiting was at least two hours, the area itself became grid locked.

And it almost seems worthless when you think of the scale of the issue. How many coaches would you need?

If one coach starts at East Croydon - the roads around which had quickly become grid lock due to people trying to pick others up due to the disruption - and then crawls through south London.

[I like many in Brighton have been to Croydon for one reason, Ikea. And I can say getting from the M25 up to Croydon is a very slow process... and here we are talking about rush hour. ]

Once through the London traffic and on to the M23 it's south to Redhill, where many trains from the south were terminating. How long would that trip take? 1 hour 30 mins? 3 hour return trip (including loading/unloading etc)? Lets go with that.

And let's say it's one coach per train carriage, we would ideally need enough coaches to match the number of carriages passing through east croydon in three hours (the time taken for the first coaches to return to make another trip).

Most trains are 12 carriages long. A very rough guess at a normal level of trains....

4 trains per hour ( 8 carriages) - First Capital Connect

4 trains per hour (10 carriages) - Gatwick Express

3 tph (8-12 carriages) - Coastway West, littlehampton etc.

2 tph (12 car) - Brighton

2 tph (8 car) Coastway East, lewes etc

That's just the coastal routes, ignores the Oxted line, South London lines, Horsham line, Redhill, etc - and to be honest I didn't check the exact numbers, just checked a couple, but if anything was conservative.

Multiple by three (hours) and using these somewhat random numbers that makes 426 coaches required!

They managed to re-open one pair of tracks around 7pm, and perhaps the emergency timetable gives an idea of how busy it is, this is for one pair of tracks, and this is just the Southern trains, there's another London Bridge train just off the screenshot, and then there is the First Capital Connect as well.

[screenshot of www.southernrailway.com (c) Southern Railway]

The point is that the railway must be one of the most inflexible systems in the world. Both in terms of transport and more generally. Even when there are other routes, both the trains and the drivers must be cleared to use them, so diversions are never simple. And unlike the roads, where a bus driver can make a common sense detour, a train needs a smal army of people to change the route, e.g. if one signaller sends the train a different way, then signallers further along need to know what is happening.

In this case not much more could be done.

The message was sent out to use Waterloo->Portsmouth for West Sussex and Charring Cross -> Hastings for East Sussex (the closest two lines either side not using East Croydon). Those for Gatwick were recommended to take a train to Guildford and then get on a small stopping service that runs to Gatwick from there, if only a small percentage of travellers got the message the route would still be packed (and so would the others, remember that peak times on normal days have people standing).

What about the future?

This must be the worst case scenario. East Croydon is the crunch point there just aren't really any other options.

I wondered if they could redirect some trains from Clapham Junction, to Guildford and then run along the line to Gatwick, but then remember this line isn't electrified, a rarity around here, and no good to the electric trains that are in use.

I thought about the line which runs through Oxted to East Grinstead and Uckfield, and the campaign to reopen the 'gap' to Lewes. This would provide a vital path down to the coast. But again some of it is not electrified, some of it is single track, plus it is quite a slow and indirect route. And of course, it would not have helped this evening as it too goes through the affected area (had the mudslide happened about a mile further south it might have been ok).

However this does argue that perhaps electrifying some of these 'gaps' around a per-electric network would help in these times.

I notice there is a line running down to Horsham which avoids East Croydon (see above) and I think this normally runs a stopping service through South London and Surrey, I've no idea if more could have been made of this line.

There's also a local line running from Tonbridge to Redhill, which perhaps could have been utilised more at times like this (e.g. for someone going to Haywards Heath, Charring Cross -> Tonbridge -> Redhill -> Haywards Heath).

I also think this is one of the arguments for HS2. It adds a level of redundancy in to the network, with HS2 and the WCML covering some of the major cities in that part of the UK.

Ultimately perhaps we need to treat essential train lines like computer servers, and try and avoid potential problems from ever happening. That may mean the water company being forced to run extra checks on pipes near mainlines, and maintain them to high quality standards, and with the same applied to other utilities and potential hazards. (of course, I'm sure such measures are already undertaken to an extent)

I wasn't affected by the problems tonight. But if there is one thing I have seen time and again it's lack of communication to/from staff at stations (and on trains) about what is happening. Both those trying to help not knowing the up-to-date information, and those staff who just don't make an effort to communicate what they do know (there have been times as I have watched a member of staff wait for each person to individually ask the same question rather than try to communicate to a group or via the public address system).

Communicating at times like this is always difficult. Though it shouldn't be impossible, even a website to refer to from hand held devices (which they use in The Netherlands) could work. Secondly, there seems to be a huge different in the quality of staff between the Sussex station (who are nearly all uniformly brilliant) to the Victoria and East Croydon staff. There always seems to be a quarter the number of staff at Victoria to that of Brighton, why? And those that are there do not want to say more than one word answers. East Croydon isn't much better. More training, and, perhaps, more pay to attract better staff in the expensive London area.

Jul 18

SPARQL query

This searches the (small subset of) British National Bibliography Sparql endpoint, which has been made available by the British Library. Much credit to them for their vision.

It shows the book titles written by a given author, where surname, and firstname are given.

SELECT ?title WHERE {
# first find the author id
?authorid <http://xmlns.com/foaf/0.1/family_name> 'Briggs' .
# the above will find lots of matches, restrict it to just Asa Briggs
# as the triple contains more than just 'Asa' in the literal, we need to use
# as filter (essentially '*asa*'
?authorid <http://www.w3.org/2000/01/rdf-schema#label> ?name .
FILTER (regex(?name, "Asa")) .

# Now find book id's which have an author with the authorid above
?bookid  <http://purl.org/dc/terms/creator> ?authorid .
# now find titles with those book ids...
?bookid <http://purl.org/dc/terms/title> ?title
}

You can try it out here:

http://bnb.data.bl.uk/sparql?query=SELECT+%3Ftitle+WHERE+{+%0D%0A%23+first+find+the+author+id%0D%0A%3Fauthorid+%3Chttp%3A%2F%2Fxmlns.com%2Ffoaf%2F0.1%2Ffamily_name%3E+%27Briggs%27+.%0D%0A%23+the+above+will+find+lots+of+matches%2C+restrict+it+to+just+Asa+Briggs++%0D%0A%23+as+the+triple+contains+more+than+just+%27Asa%27+in+the+literal%2C+we+need+to+use%0D%0A%23+as+filter+%28essentially+%27*asa*%27+%0D%0A%3Fauthorid+%3Chttp%3A%2F%2Fwww.w3.org%2F2000%2F01%2Frdf-schema%23label%3E+%3Fname+.%0D%0AFILTER+%28regex%28%3Fname%2C+%22Asa%22%29%29+.%0D%0A%0D%0A%23+Now+find+book+id%27s+which+have+an+author+with+the+authorid+above++%0D%0A%3Fbookid++%3Chttp%3A%2F%2Fpurl.org%2Fdc%2Fterms%2Fcreator%3E+%3Fauthorid+.%0D%0A%23+now+find+titles+with+those+book+ids...%0D%0A%3Fbookid+%3Chttp%3A%2F%2Fpurl.org%2Fdc%2Fterms%2Ftitle%3E+%3Ftitle++%0D%0A++%0D%0A+++++++++++++++}

or paste it in to:

http://bnb.data.bl.uk/sparql

It's also available via Kasabi (beta) from Talis

http://labs.kasabi.com/explorer/sparql/sparql-endpoint-british-national-bibliography-bnb

I'm noting it here as it's the first real SPARQL query I've made myself.

If you think you can improve it, or have any comments please do note them below, or via twitter.

Jul 10

News of the World

On thursday I left a meeting that had just come to an end, I sat at my desk and a minute or so later some one else who had been in the meeting popped their head round my door. "Have you heard The News of the World is closing this week?"

It's not often a piece of news like this leaves me utterly shocked. The NotW had, quite rightly, become the nation's punch bag. But not for one second had I seen this coming, it was unthinkable. From what I can tell the NotW staff themselves had only heard about it 30 mins (or less) earlier, the earliest mention on my twitterstream was 12 mins a go

http://twitter.com/#!/chriskeene/status/88999219351928832

If you read one article about it then make it this excellent New York Times piece http://www.nytimes.com/2011/07/10/world/europe/10britain.html?_r=2

This Economist article, written before it was announce NotW would close is also worth a read http://www.economist.com/node/18928406?story_id=18928406

The move was canny by Murdoch, especially with the BSkyB deal being the Golden Goose. I dislike the man, but do respect his for making nearly a life time of business savvy decisions.

Perhaps my biggest worry is the long term power News International has had over our politicians and governments. An MP who criticises the Murdoch empire (or suggestions something Murdoch does not like) could receive a phone call threatening blanket negative stories. I was fairly young when David Mellor was caught up in a scandal - an affair - but I wasn't aware he had put forward proposals as Minister for the Media, which had upset Murdoch and co. There have been suggestions that Gordon Brown was told to 'shut up' Tom Watson regarding phone hacking.

When Cameron met Murdoch for the first time, aides of Murdoch was shocked and angry, they had told him how to answer Murdoch's questions, but he had decided to use his own answers. He had to work hard after that to make up for such a mis-step.

The grey lines between the press, Government, and the police, and the way they all have questions to answer in this mess in worrying.

If this week has made Murdoch's hand weaker then good.

I also want to touch on the 200 journalists being made redundant. I've seen more tweets showing sympathy for them,and disgust that people are gloating over their loss, then i have tweets that do indeed gloat over their loss.

200 people loosing their jobs is never good. But this week the staff of Habitat have a similar fate, so too do the thousands of workers at the Bombardier Rail engineering plant in Derby.

Many have made the point that these people are innocent and the illegal activity was carried out by a previous generation of staff.

This is true. But the NotW is well known for inciting hatred towards minorities, making single mothers seem like the devil, and showing immigrants and the unemployed in the lowest of light, often using pure lies. Most on the paper had nothing to do with this. but this was not a team that was whiter then white. They may not have done anything illegal but they were producing week after week copy that was morally and ethically dubious. Innocent people was misrepresented, and hounded on their door step, for a story. For some of these people, I will morn their job loss a little less than the many others who loose theirs each week. sorry.

I should add I didn't see anyone gloat. I saw people celebrate the closure of a paper, which yes by implication leads to job losses. But then we celebrated the demise of Big Brother, no one commented on the loss of media jobs. So yes, many people (including me) do care about their loss of work, but we can also celebrate the end of something we thought did no good to our society.

I've seen a number of comments and posts blaming the BBC and Guardian the closure. This is nonsense the blame lies with those in News International who decided to close the paper, and let it happen in the first place. I find their mis-placed anger bizarre.

Finally, like many, I find it impossible to see how an Editor was blind at all this going on. If they were in the dark they must be the most incompetent manager imaginable. In any case they created a culture where this was seen as ok, and in any case they - Rebekah Brooks - need to go.

It seems clear that there is probably more to uncover, both at the NotW and other papers. The PCC is useless and needs replacing, with something with teeth. Journalists will claim this will be the end of the world. It wont be. TV news is covered by OFCOM, I do not see Newsnight avoiding difficult questions because of it.

Jul 3

Inside Google+ — How the Search Giant Plans to Go Social | Epicenter | Wired.com

The parts announced Tuesday represent only a portion of Google’s plans. In an approach the company refers to as “rolling thunder,” Google has been quietly been pushing out pieces of its ambitious social strategy — there are well over 100 launches on its calendar. When some launches were greeted by yawns, the Emerald Sea team leaders weren’t ruffled at all — lack of drama is part of the plan. Google has consciously refrained from contextualizing those products into its overall strategy.

via wired.com

Great article about Google+ and how it is part of a larger move towards social. Larry Page moved his office to be in the same building as the Social project

May 29

What Happened to Air France Flight 447?

via nytimes.com

Great article with background about Air France flight 447. An example of how good journalism can be.

(updated information after this article went to press can be found here http://www.bbc.co.uk/news/world-europe-13572569 )

May 29

Academic publishing: Of goats and headaches | The Economist

via economist.com

"HOW much would you pay for an annual subscription to Small Ruminant Research, Queueing Systems or Headache? University librarians pay rather a lot. In Britain, 65% of the money spent on content in academic libraries goes on journals, up from a little more than half ten years ago. With budgets tight, librarians are trying to resist price increases. But Derk Haank, the chief executive of Springer, a big publisher, is firm: “We have to make a living as well.”

And what a living it is. Academic journals generally get their articles for nothing and may pay little to editors and peer reviewers. They sell to the very universities that provide that cheap labour. As other media falter, academic publishers have soared. Elsevier, the biggest publisher of journals with almost 2,000 titles, cruised through the recession. Last year it made £724m ($1.1 billion) on revenues of £2 billion—an operating-profit margin of 36%."

Good general article about the state of academic scholarly publishing (I'm one of the commenters but my inadvertently does not show). And yes, the Library I work at does have access to it: http://xd5.be/2p )

May 19

Communicating knowledge: how and why researchers publish and disseminate their findings | Research Information Network

via rin.ac.uk

From Communicating knowledge bibliometric analysis

Key findings
Dissemination practice
• There were significantly more outputs per author in 2008 than in 2003, particularly in
Biomedicine, and social sciences
• There were significantly more journal articles, and fewer monographs, in 2008 than in 2003
• There were significantly more multi-authored works in 2008 than in 2003, particularly in
social sciences and physical sciences
• There were significantly more inter-institutional collaborations, and more international
collaborations, in 2008 than in 2003
Citation practice
• There was no difference in the average number of citations per output between 2003 and
2008 overall
• There were significant differences between disciplines in the numbers of citations per
publication – humanities cite the greatest number of works on average; engineering the
fewest
• Monographs had an average of over 230 references each, compared to 38 for journal
articles and 47 for book chapters
• Significantly more journal articles, and fewer books and grey literature works were cited in
2008 than in 2003
• Biomedicine, physical sciences and social sciences cite twice as many articles per
publication as other disciplines
• Humanities, and, to a lesser extent, social sciences and education, cite more books per
output on average than other disciplines
• Social sciences, and, to a lesser extent, education, cite more grey literature per output on
average than other disciplines
• Social sciences, education and humanities cite more websites than Biomedicine, physical
sciences or engineering
• Books and book chapters are most likely to cite books/book chapters
• Conference outputs are most likely to be cited in conference proceedings

About