RDFa actually became a mandatory requirement for all central government consultations published after 1st January 2010, but in the past year we have seen very little use of this new wealth of freely available data. There were rumours that Directgov planned to use it to import consultations, but I can’t find any evidence of this on their site.

We were sad that we’d gone to all the trouble* to incorporate RDFa into our clients’ consultations, and nobody was making use of it, so we decided to do something with it ourselves. We’ve made an Aggregator that can collect together consultations from any website as long as it publishes its consultations in an RSS feed and includes the appropriate RDFa markup. It also publishes its own RSS feed so that the aggregated consultations can be fed into a further tier of applications.

We’d be really interested to hear of other apps that are making use of RSS and RDFa in relation to consultation data. Surely we’re not the only ones?

*actually it wasn’t much trouble at all thanks to the helpful guidelines from the COI.

Andy’s post is a philosophy on how to make “good” apps. It’s a great post on the principle of having 80% practicality, 10% glamour and 10% character. Ideologically, this will provide you with an app that people will love and make you a multi-millionaire! However, it doesn’t always work out that way. We’ve seen it many times on Dragon’s Den where a young, hopeful entrepreneur presents their idea, only for the dragons to rip them apart and leave them empty-handed with their dreams in tatters……So who says it’s “good” – my argument is stress the importance of user-centred design.

Who holds the purse strings? Your wife, your boss, the queen? I work in part of a team that develop large scale websites for government organisations as well as advertisers with large budgets hoping to attract millions. The app / pitch can sometimes appear to be king. It’s what wins the client over and wins us contracts. However, that doesn’t always define your app as “good”,  just because the CEO of the company loves your app doesn’t mean Joe Bloggs who subscribes monthly and uses your app day-in day-out will too. If Joe Bloggs and countless others like him, hate your app and it flops……is your app still “good”.

Andy’s model sits perfectly in terms of assessing the values of the user. Ultimately, an app needs to work – 80% practicality. Too often products are thrown by the way side for not solving a problem or doing the job it was meant to do. This is particularly emphasised in our consumer culture today. The user’s value may indeed fluctuate between glamour/character and practicality as good marketing is always effective in blurring a user’s sense of need.

For an app to succeed, the user’s voice is priceless. An app will either thrive or dive by the user’s voice. This can be seen in Apple’s App Store. Angry Birds is currently no. 1 paid for app. This follows Andy’s model of 80% practicality – it’s essentially a great game. It’s engaging, not to difficult, but challenging enough to leave you wanting more. 10% glamour – it looks good, but more importantly it doesn’t distract from the game. The graphics don’t slow the game or make things difficult to see. 10% character – the birds are fun. There are talks of a TV series based on the strength of the characters in the game.

The user ratings and reviews for Angry Birds has propelled the app to the top of the store where it has sat for a good number of months. When making a transaction decision, advocacy is key. A recommendation from a friend, a high rating or positive feedback can carry a lot of weight for a user in whether to take the plunge with your app. Andy’s model is the foundation for creating a “good” app but ultimately the end user will decide whether the app is indeed good.

Hopefully, you’ll see the importance of valuing the user in every stage of the development of an app. User-centred design starts and ends at the user. It continually comes back to the issue of “who is this for?”, “what problem are we solving” etc. it uses usability testing to measure how we’re doing in the process, whether we’re still on track or veered way off course. It isn’t a launch and cross fingers….

The chart below shows the share of tweets by political party, split by question over the debate.

Data provided by Delib and The Viral Ad Network

[update]

News just in (straight after this post was published) – YouGov poll suggests that the Conservatives won the debate – which coincides with the data we recorded above.

[update]
1. Gez thinks these results are massively skewed by a joke from Simon Pegg.
2. Twitter is all up in arms about YouGov CEO standing as Tory candiate.

You can find it here: http://www.delib.co.uk/leadersdebate/

On a separate note, Google Analytics now provides an official way to insert their tracking code asynchronously, which is really cool.

http://code.google.com/apis/analytics/docs/tracking/asyncTracking.html

We’ve built (and are continuing to iteratively improve) Citizen Space, an open-source online consultation management and creation system designed primarily for local authorities and other governing bodies.  It’s currently being used by Bristol City Council, Sutton Borough Council and a large partnership led by Cumbria County Council, with more rollouts hopefully coming soon.

Through Citizen Space, organisations can store, organise and publish details of all their consultations – hopefully, exactly the kind of information that it’d be really useful to be able to access from a variety of sources, to include in apps and to run analyses/queries on.

As you probably know, that’s what Linked Data is designed to facilitate; it’s a way of marking up web pages semantically so that software (like browsers, search spiders etc) can understand more about the content and so that it’s easier for citizens to ‘cut and combine’ the information in ways that are relevant to them.  The Government’s into it in a big way: all central government consultations need to include this markup as of the start of 2010 and DirectGov already uses it to pull consultation details into its listings.

We’re currently making a few changes to the code of Citizen Space so that all consultation details entered using the system are marked up as Linked Data – using RDFa, to be precise.  This means that all consultations inputted into Citizen Space will include this additional metadata, making the information therein more readily and widely available.  So, for example, DirectGov could pick up all of a local authority’s consultation details and pull them directly into the ‘local consultations’ part of their site, or someone could make an app to compare consultation activity with other data, such as population or earnings surveys.  That’s a lot of potential added value for something that’s really not that hard to do – and it won’t affect the appearance or user experience of Citizen Space for end users or administrators as the data will be taken from existing fields.

It might also be worth mentioning that this is in addition to the existing RSS functionality built into Citizen Space – for example, you can create an RSS feed from any search query in Consultation Finder, meaning you can take the content and embed it into a blog or community website.

Basically, we like sharing and we like making the most of data – so we reckon RDFa’s a good thing all round and an easy win in many ways.

Pain can be a beautiful thing. And in my case quite an inspiring thing too. For the last 5 days I’ve been stuck in hospital with severe back pains – which have meant I’ve been totally bed bound, and completely dependent on the nurses at Kings College Hospital London (who have – by the way – been really great!)

Anyway, lying in bed – in pain – inspired me yesterday morning: could I be of any use to hospital lying here? The hospital staff are being great – but could I add anything – could I help improve how the hospital is run?

And the answer is of course YES. As a patient I – and the hundreds of other patients – are in the best place to help the hospital improve, as we’re the ones at the coal-face, experiencing the end products that the hospital delivers – whether that’s a quick fix in A&E, or a longer stay in one of the wards.

The only issue is how you collate and make sense of this collective experience. And the answer to that is the internet – create a crowd-sourcing website with a focus on collating “ideas of how to improve the hospital” and let patients share ideas, and self-organise the value of those ideas via rating and commenting systems. Given I’m the co-founder of an e-democracy company – Delib – that specialises in citizen empowerment – this was the easy bit!

So lying in my hospital bed – I asked Andy, Jess and Dave in Bristol to quickly put together a patient crowd-sourcing site using our one of our apps – and 2 hours later we launched “Help us Improve Kings”.

Check the site here (and add ideas if you have any!)

With the prototype site up and live, I’m now in the process of getting patients to take part and share their experiences and ideas – a bit of a tricky feat given I can’t walk, but they’re coming in slowly as I lynch people walking past my room! I’m also in the process of talking to the Patient Involvement team @ Kings – as obviously to make this work, we really need them on board to actually turn the ideas generated into concrete actions.

So there you go – an example of bottom-up patient power – empowered by the wonders of broadband and a wonderful set of e-democracy tools. Social media empowerment at its finest!

And most importantly this is a lesson to all those brands / businesses / government departments out there who are trying to work out how to make their business work better – the answer: empower your customers and employees and they will in turn help power your business!

we like this !

The Creative search box with the text 'how to kill birds' in

Now you can monitor all the LOLs and WTFs to your heart’s content…

Thanks to Mark Pack

To summarise, I feel that focusing on twitter ends up creating a very bad sample for any kind of opinion research, practically ignoring the effect of Facebook, Beebo, Myspace, Youtube, Search engines, News sites, Email,  Blogs, Forums, Instant messaging, and all the millions of other websites on-line.

What is more, I believe that focusing on twitter so strongly is what throws twitter’s collective opinion out of line with the rest of the internet. Online marketers going to twitter to measure internet opinion is like a market research team only inviting people who work for market research companies to give feedback on a product.

Relative Traffic

The first release that caught my eye points out that Facebook now accounts for 14.5% of all UK page views.  That means that more than one in seven views in the UK is on Facebook.

In fact, Robin Goad, Director of Research at Hitwise, says “During September, Facebook received more page views than Google UK, eBay UK and YouTube combined”.

Defenders of the practice of focusing on twitter will point out that the value of comparing page views has decreased massively in the last ten years. This is true, but Facebook is notable because viewers do spend a significant amount of time viewing their site.

As Robin Goad continues, “[Facebook's] average visit time increased from 19 minutes 59 seconds in September 2008 to 26 minutes 14 seconds during September 2009″. That’s almost half an hour per visit to the social networking site!

Twitter users

Again, people may defend twitter at this point by saying that (although people don’t spend much time on it on average, and people don’t visit it very much on average), it offers a good cross section.

If twitter did offer a good cross-section of the internet public, I wouldn’t be writing this post – online opinion research is all about collecting a good cross-section of people, not about finding every single comment.

Facebook has over 49% of the social networking market in the UK, where twitter has a 2% share.

How do I know that that 2% is not representative of the UK? At this point I’ll turn to Alexa. Alexa is the traffic measuring service that’s well know for having a very strong bias towards online marketers. The reason being that it gets it’s traffic data from people who install the Alexa toolbar – and the main reason you would install it is if you are a web marketer.

Alexa has such a strong bias, that it once reported the personal blog of Matt Cutts (head of web spam at Google) as being more popular than Yahoo!.

Recieving a much higher ranking from Alexa than from other measurement services is a sign that a website’s viewers are largely online marketers themselves. Let’s see how twitter compares

According to Hitwise, twitter is the 38th most visited site in the UK. According to Alexa, twitter is the 12th most visited site in the UK. That’s more than enough evidence in my opinion.

“Influencers”

Another well used defence of the over-reliance on twitter is that (since twitter doesn’t have a good cross-section of users, involves 25 times less people than Facebook, and people don’t spend nearly as much time on it), those people must be really important. In fact, they must be “influencers”.  (I feel people’s logic works backwards here – starting at the assumption that the over-use of twitter is justified, and trying to find the justification).

For some reason, people keep pulling out the “influencer” card, even through study after study has shown us that “influencers” are at best impossible to detect before the event, and at worst wouldn’t have any effect even if you found them (See the 2007 Harvard Business Review list).

The argument for influencers put forward in Malcolm Gladwell’s “the tipping point” was well and truly shot down (both in acedemia and in business) several years ago, but for some reason it’s still used all the time when defending the over-use of twitter.

Direct Marketing

So perhaps the final possible defense of focusing on twitter so much is that it achieves greate direct marketing results? After all, Dell claim to have made millions of dollars through twitter. Sadly (as you probably expected as you’ve got this far through this post), this isn’t the case – Twitter sends under 10% of it’s outgoing traffic to transactional websites (c.f. 30% for Google, and 15% for Facebook). In fact, “During May, Google UK sent 365 times more traffic to transactional websites than Twitter. “.

Going Viral with Twitter

Twitter did appear to have a significant viral effect on the traffic of some sites. In particular, twitter has sent a huge amount of traffic to  twitpic.com – so it seems that twitter does have the possibility to make some sites go viral – so long as your site is dedicated to twitter (and preferably has “twit” in the title)

Of course, the ironic thing is that if the people creating social media analysis tools are using their own tools, they won’t pay much attention to this post – simply tweeting  “twitter is overrated” would have been taken as purveying more of an opinion.

Tim Wintle

Tasters

First the notes from a 2008 talk given by Lillan Lee from Cornell University. Lee’s topic is “…the flood of interest in: sentiment analysis, opinion mining, and the computational treatment of subjective language.”

This is a good ‘who, what, why, how’, featuring:

• - background
• - useful stats
• - an exploration of the broader implications of sentiment analysis in politics and business
• - a basic outline of the scientific problems in classifying sentiment.

Read Lee’s piece here.

Meanwhile, the epicly-named Dirk Shaw has a blog post from July 2009 asking Sentiment analysis, How much is good enough?. It’s a short post explaining the basic differences between manual and automated sentiment analysis.

Thinking Deeper

In a post from June 2009, Irfan Kamal describes the approaches Ogilvy PR are taking to make sentiment analysis work for clients.

The Ogilvy post is useful in one respect because it shows the debate is still fledgling and the possibilities are wide open.

In another respect it helps illustrate that the people-or-software-or-both question of classifying sentiment is only one dimension of the problem. Sentiment analysis has to actually be of value to the people using the results.

In my view, the most useful output would be actionable insight, i.e. information that is directly useful for making decisions. That comes with a health-warning though: if we’re to make decisions on the basis of data, we had better be sure the data is completely valid, and leads to accurate conclusions.

It’s not clear yet that sentiment analysis will be able to deliver in such a concrete way, and at the level, for example, of government policy, sentiment data will find a place, but should be treated with healthy caution.

Marta Strickland has a post on these issues from September 2009. Focussing on product reviews, she identifies Five Reasons Sentiment Analysis Won’t Ever Be Enough, and concludes “What are we really trying to decide with this data? And are we asking the right questions?”

Wrap Up

So it it hopeless then? Is sentiment analysis a turkey? Absolutely not. Asking critical questions absolutely indicates that smart people are taking sentiment analysis seriously, and figuring out how best to do it and how best to use it.

Team Rubber has a working model for how we want to use sentiment analysis – for both our advertising and policy-making work. We’re evolving the model on the basis of trials, and we’re asking ourselves hard questions about how sentiment analysis can be valuable for our clients.

Enough about that (but more in future). Meanwhile, I’ll wrap up with a piece from each of the manual and automated perspectives. Nathan Gilliat has a September 2009 post on Scaling Human Analysis, while an April 2008 paper from Google discusses software approaches for Building a Sentiment Summarizer. Enjoy.