Guardian Crime Maps

The Guardian have teamed up with Spatial Key to create a variation on the recently announced UK street level maps (http://www.police.uk). See here for a look at the Guardian crime maps 
(http://www.guardian.co.uk/news/datablog/interactive/2011/feb/15/uk-crime-map-interactive-tool). They seem to have forgotten about Scotland though as this image highlights! Unless of course there are no violent crimes in EH1? Aha but the reason is that the crime data (from Police.uk) only covers England and Wales. What do you think – do you like the Guardian site? It offers a nice way of comparing 2 locations at once but as others have mentioned before – is this data meaningful?

GeoMob Event – Feb 2011

I went to the BCS (British Computer Society – http://www.bcs.org/) on Thursday as they were hosting the quarterly GeoMob (http://geomobldn.org/) event. I got there slightly late so missed the food and drink! I did manage to meet my geo-pals Rollo and Steven though which was good.

The event itself quickly got started and the first talk was by Dan Pett from The British Museum. Of the evening, his was the most ‘traditional’ GIS-like. He talked about his creation of http://finds.org.uk (a website to record archaeological finds) and the importance of spatial. You can imagine that knowing where someone found some old coins is pretty useful for discovering the history of the area. Technology wise Dan is using YQL (Yahoo Query Language) and Yahoo GeoPlanet WOEID (Where on Earth IDs) to locate where people have recorded finds on the ground. The WOEID allows you to reverse geocode results. Take Crofters Farm, Berkshire and turn into 51.48631, -0.78941. Most of the positions recorded on the finds website are degraded to prevent opportunist metal detectorists from looting a site! The finds.org.uk site uses old Ordnance Survey maps which are served by the National Library of Scotland – http://maps.nls.uk/. The presentation was very interesting and the next plans include creating a database of ancient placenames – something already started by the Pleiades project.

The 2nd presentation was by Simon from Facebook who works on the Platform Engineering team in the UK. Facebook now receives over 250 million hits every day from mobile phone users and has another 250 million regular desktop users. There are 1 million regular developers. These are phenomonal numbers and highlight that the popularity of facebook is showing no signs of waning. Last week in the UK Facebook launched Deals and Places. Simon also showed the Graph API which allows access to all the Facebook data (well provided you have the right authentication). The question of privicy came up from the audience and was answered pretty well. But the question still remains as the mobile facebook site does not have any privacy setting options. Of those 250m how many realise how much of their information is private and how much is public? The site http://openbook.org highlights the privacy issue by showing all the details of people who haven’t secured their profiles!….a little odd. Anyway on the plus side for Facebook mobile users Facebook connect allows single sign on (you only have to type your user name and password once) to a number of sites. Not sure what would happen if someone stole your mobile though – would they then be able to go on a spending spree? Unfortunately there wasn’t time to go into HTML5 (many of the audience were dissapointed). Simon (from FB) suggested that for getting apps onto iPhone etc. a wrapper could be put onto the app store which contained a HTML5 application. This would be much easier to maintain for mobile developers. At the moment mobile developers havce to develop for Android, Apple iOS, and others. With HTML5 it would be (possibly) one size fits all.

The 3rd presentation (well talk as no there were no slides this time) was about a site called http://www.likeourselves.com. Pardeep is the CEO and wants to break down communication barriers with this site which will ‘help people meet people’. The site allows people to meet who share same interests. People who like snowboarding form a large group, say they want to meet up and then choose who they want to talk to by sending messages. But the main difference is that if you send a message the other person won’t know until they send you a message. Apparently this will mean that you’ll feel better as you haven’t been rejected. It’s explained as such on the website…..

The biggest problem in meeting someone is the final barrier of walking over to chat. It’s not a problem anymore. The ‘hi’ feature clears that final barrier of starting a conversation by anonymously confirming that you want to talk to each other. It’s like whispering to a 3rd person, which person you want to talk to and the 3rd person only tells you both if you match. In this case, you can ‘hi’ each other before you even leave the house, chat online and then walk into a place that feels friendly. Likeourselves is not a dating site, it simply makes meeting new people refreshing and comfortable.”

But having said all this does it not take does it not take away the feeling of chance? Do I really want to go out somewhere new and know who I will be talking to (even though I’ve never met them). Technically the likeourselves site is built on google app engine with Google maps API v3 and google local. They’ve got 3500 users currently. Will it take off? Watch this space!

The 4th presentation was from Ben Dowling of Geomium (http://www.geomium.com). This was a technical presentation and went into what you might use to build a simple (mobile) geoapp. WOEIDs were mentioned again and reverse geocoding. Gowalla was mentioned – I’d never heard of it but there is a good explanation here http://en.wikipedia.org/wiki/Gowalla . Privacy was again mentioned with the site http://www.pleaserobme.com mentioned (along the lines of OpenBook). The use of HTML5 was mentioned too as the future and hopefully a way around having to develop against many different mobile platforms.

The final presentation was from the CEO of Vouchacha (http://www.vouchacha.com), Ben Brown. He used to work for Lloyds Banking and Affiliatewindow and set up Vouchacha 2 years ago. This is a effectively a digital voucher company (similar to Groupon, and Geomium and the new Facebook Deals). The idea is to target users within walking distance of their current location. An example was given about the archetypal coffee shop and targeting the 20 nearest customers with a deal. 5 years ago this was all called Location Based Services but now there are companies making serious money (like Groupon). Ben said that by 2014 there would be an estimated 300 million mobile coupon users. Also in a study of location based apps it was found that coupons were the 2nd most useful feature that customers wanted. The first being GPS (as I imagine it would be difficult to use location based apps if your phone didn’t know where you were!). Ben is also looking to the future with Augmented Reality (I, like many, predicted the rise of AR in my Geospatial Trends 2011 post). Technically Vouchacha use cloud based services to allow scalability. If they can battle through all the digital voucher services which already exist then Vouchacha may be going places. They’ve signed an agreement with myvouchercodes.co.uk (one of the UK’s bigger voucher websites) which surely gives them a good start.

In conclusion the Geomob event was worth attending. I learned a lot but 4 out of 5 presentations were mobile rather than geospatial (no bad thing and it gave the evening a bit of a twist). I’ll make sure I keep up to date with what’s going on at future Geomob events. I think for traditional GIS / Geospatial people this event was a bit too developer-skewed, but there was much to be interested by…..and not a single mention of E S R I!

 

Auto Detect Roads – Bing Maps

As Steve Coast (OSM guru) moved to Microsoft a few months back more collaboration between Bing and OSM was to be expected. Word* came my way of an application developed by Steve Coast which allows a user to click 2 points on a road on a Bing Map satellite image. The tool then automatically follows the road and creates an OpenStreetMap XML file of the route. You could then upload your XML to the OpenStreetMap database I guess.

I experimented a bit with this and results were varied! As the autodetect algorithm relies on colour to decide where to place the next road node there are problems when the colour of the road in the image is the same as the surroundings…..as in the image below.

I did a bit of computer vision in my MSc (on image recognition of street furniture!) and I came across similar problems as these. Image recognition techniques based on colour are always more error-prone. But it’s not all bad. The example above isn’t really fair. To the human eye it is difficult to see the road between the hedges and trees in the centre of the aerial image. I ran the application on a wider road in the countryside and the results were pretty impressive. Also in semi-urban environments the results were good.

This application is specialist but should save OSMers time in collecting and digitising vector road data. As for attribution….well that’s another story!

*Thanks to Nick for this information.

 

Police.UK launch

 

So this week saw the launch of http://police.uk and it being a web mapping site I had to share my thoughts on it. Basically it’s a crime mapping site and provides  crime statistics down to street-level for the whole of the UK.

Most of the news stories this week have been about the fact that the site crashed due to over-user. Some 75,000 hits a minute according to the BBC(http://www.bbc.co.uk/news/uk-12336381). It’s been the same recently with many other government sites when data is opened up. When the Ordnance Survey (http://www.ordnancesurvey.co.uk/opendata) released their OpenData iniative their site was down for days.

Other stories have been about the anomalies caused by the site which have given rise to incorrect “highest-crime” areas. This could be down to 2 factors:
1. crime stats being aggregated for streets with less than 12 properties
2. crime stats being unrealistic as they’re based on police recorded crime (http://www.guardian.co.uk/commentisfree/2011/feb/03/crime-map-information-theresa-may). Really the crime information on the site is not as detailed as you’d imagine (despite all the media furore).

The financial cost has been mentioned by many. Rock Kitchen Harris (RKH) were paid £300k for the site. A web mapping geek (not me!) peeved off with the amount the UK Home Office paid has set up an alternative site here http://www.crimesearch.co.uk/ (not as pretty but still effective. Not sure I agree that much with the cost argument in The Register (http://www.theregister.co.uk/2011/02/04/crime_mapping_police_uk/page1.html). The PR the UK Government got from the launch of police.uk (news all over the UK papers and headline news on BBC, ITV, C4 and others) has surely made the amount the the Home Office paid for the development worth it. The main reason for the site is Government PR surely. The more sensible option (lower-cost) should have been to publish the police data on the http://data.gov.uk/ website. Perhaps there will be a release of ongoing monthly police data site on the data.gov site?

RKH (http://www.rkh.co.uk/) have released an API for the new police.uk site – see http://policeapi2.rkh.co.uk/api/docs/. I’ve signed up to this and will start to publish some custom web maps on this site….watch this space!

Elsewhere around the world there have been crime maps for quite a while, most notably in the US. In the US http://www.spotcrime.com/#90210 has pushpins for crimes in a particular area and the ability to select dates and types of crime. So now we have police.uk. What does it tell us? Is it useful? Who will gain from the site? Let me know what you think below!