The following is part of a chapter for a forthcoming book on online journalism. Contributions welcome.

Maps have become a familiar part of the news language online due to a number of advantages:

• They provide an easy way to grasp a story at a glance
• They allow users to drill down to relevant information local to them very quickly
• Maps can be created very easily, and added to relatively easily by non-journalists
• Maps draw on structured data, making them a very useful way to present data such as schools tables, crime statistics or petrol prices
• They can be automated, updating in response to real-time information

News organisations have used maps in a number of ways:

Mapping public data

In the US there has been a rich history of mapping crime statistics online, stimulated enormously by Adrian Holovaty’s ChicagoCrime.org, which he later developed into EveryBlock, a website which doesn’t just map crime – but also planning and liquor applications, filming, news stories, street closures and restaurant inspections. EveryBlock pulls in information from a range of sources and displays the results based on the zip code you enter, giving you a picture of everything happening local to you.

In the UK the most significant mapping of public data has been around elections. The Telegraph election maps, for example, pull from a database to provide links to specific statistics and reports. The BBC, meanwhile, have provided maps that you could change based on your own prediction of the ‘swing’ a political party might experience.

More recently, however, as there has been an increasing move towards publishing public data and increasing use of the Freedom of Information Act to obtain public data, types of data have broadened.

Working with the public

When a number of motorists started reported problems with their cars, the BBC was able to gather data from viewers and website visitors to compile a map of cases richer than any motoring organisation or transport department. That data allowed them to pinpoint the particular petrol stations where contaminated fuel was being supplied. During floods in Berkshire BBC Radio Berkshire also used maps to show the worst affected areas, and what people were saying, along with data about emergency services.

On a regional level the Manchester Evening News plots information about congestion and roadworks on its travel map; and the Hartlepool Mail has used its readers’ contributions to map pot holes and derelict areas of the town. On the less serious side, newspapers have used maps for sightings of an unusual bird, to identify where readers are living around the globe, and to map sightings of unidentified flying objects.

Increasingly news organisations are providing raw data to users for them to create their own maps. The Guardian, for example, provided travel expenses data for MPs for users to analyse. Tony Hirst, publisher of the blog OUseful, visualised that data on a map in a way that made it easy to tell at a glance which MPs were claiming more for travel than other MPs who lived nearby. The newspaper has a Datablog which regularly releases data for users to visualise in different ways (The Times also have one, called Times Labs Blog, which at the time of writing did not release data; the New York Times does release data in its Visualization Lab).

Maps and mashups

Maps work particularly well when combined with another service that includes geospatial data – for instance, latitude and longitude, or a postcode or placename. During the US election a number of these ‘mashups’ appeared showing, for example, Twitter tweets or YouTube videos about the election displayed on a map. During the Beijing Olympics BBC Sport used similar technology to display tweets, blog posts and photos on a map of the Olympic village.

Image maps

In addition to traditional cartographic maps, it is possible to use images of anything as a ‘map’ that you navigate in the same way with your mouse, clicking on particular areas to bring up relevant detail. Examples have included the route of a race or river, a building’s floorplan, a timeline, line or bar chart, or even a group photo. In one excellent example, NPR took a picture of health lobbyists attending the hearing of a new health bill. Users could click on individual people to find out more about them – and were also invited to identify others in the picture. Similarly, The Guardian’s Interactives section often includes diagrams where you can roll over different areas to find out information about different aspects of the process or story.

A few years ago you needed to have skills in Flash or Dreamweaver to create an image map. However, a number of web-based tools (e.g. Vuvox and FineTuna) have since been launched that allow you to create these more easily.

Geotagging and the semantic web

Both the rise in mapping and a rise in people accessing news on mobile phones has created a demand for ‘geotagged’ (or geocoded) news. Geotagging a news article means adding geographical information to it – usually, latitude and longitude – in a way that makes it easy for search engines and news distribution platforms to understand what area that news article refers to.

In practice this means that if you are on a mobile phone with GPS technology you can search for ‘restaurant reviews near me’ or ‘crime stories near me’. Likewise, if you were looking for a new house you could easily find stories about the local schools, or plans for new buildings. Many search engines take into account the searcher’s own location when bringing up search results – so including geotagging in news stories would also increase the likelihood of your content being found by a local searcher.

Most news organisations are exploring geotagging in some capacity – in many cases, changing their content management systems so that journalists can add such information when publishing a story. Some have used this information to launch ‘hyperlocal’ parts of their news websites that allow users to read stories specifically about a particular postcode.

At the same time, organisations like Reuters have developed technologies that add geolocation data to stories after they have been written – using semantic web technology to look for place names in the article text and disambiguate them, understanding that the ‘Birmingham’ referred to is the UK’s second city and not the place of the same name in Alabama, US.

The weakness of the latter approach is that it can only work on the information contained in the article, which may not be specific enough. Meanwhile, journalists geotagging their own articles need to be aware of the privacy and legal implications in, for example, identifying the specific house that a criminal lives in. Normally postcodes in the UK are vague enough for this not to happen – while licensing arrangements mean that Google Maps’ mapping of UK postcodes varies in accuracy.

How to involve users in mapping

There are 3 broad approaches to mapping contributions from users. The first is to process every contribution manually – taking emails, phonecalls, comments and texts and entering them into the map yourself. This has clear advantages in being able to verify the information and keep the map working properly, but obvious disadvantages in the amount of time it requires for a journalist and how long it takes for the map to be updated.

A second approach is to publish the map in editable format – that is, allow anyone to edit the map directly. This obviously has the advantage of not requiring any further work from the journalist other than checking the map regularly and correcting any mistakes. However, it does require a certain level of technical competence from users, and if you’re editing a Google Map, which doesn’t yet have any granularity of control, you will find users accidentally editing each others’ entries and the title of the map itself.

The third approach is to part-automate the process in a way that addresses the weaknesses above. You can, for example, set up a Google Map so that it displays data from a Google spreadsheet. Publishing that spreadsheet and allowing users to edit it will likely result in more contributions, fewer errors, and easier correction (if anyone vandalises the spreadsheet you can easily ‘revert’ to previous versions). You can also create a Google Form for that spreadsheet, which you can publish on a website for users to fill in by answering a few questions. This makes it even easier for users to enter information and prevents them editing others’ – although it may mean duplicate entries from different people entering the same information.

Have I left any considerations or concepts out? Any great examples that deserve highlighting? Let me know