Nigerian postcodes – a case for open data
For quite a while, the Nigerian postcode system has been operational but when you ask people to fill forms, they fill in 234 or something similar for their postcode. There’s very little information available for postcodes and very little projects even use them. I will attribute it majorly to ignorance.
Why is a postcode system useful? It’s essential for routing and location identification purposes. When someone specifies his street name as Adenuga Str in Lagos, is he referring to the Adenuga Str in Alausa or the one in Oremeji Ifako? With a postcode system, this ambiguity is removed because you can identify which location is being referred to. Another use is for package or mail routing or routing in general. If I need to deliver three packages to locations A, B and C. I can effeciently route those packages if I know that B comes before C and than A along a particular route. I’m talking from a logical perspective. I might not know what other industries use postcode systems to do and that brings me to the case of open data.
The open data movement is one that clamours for and encourages the sharing of data in a raw and as discrete form as possible. Why? The advantages are similar to why web applications have APIs – to enable repurposing or data reuse. If I’m using postcode data to know which hospitals are closest to which postcode locations, another project might use it for something equally or even more useful.
Postcode data (until now) has been something you could only access from a few websites – mostly the Nigerian Postal Service website and a site like AddressDoctor. The utility only enabled individuals to know their postcodes – very little utility I would say. However, with my effort and the efforts and inspiritation from folks like Dipo Fasoro of WangoNet (who’s a strong believer and evangelist of open data) and Kayode Muyibi of Nairahost, we’ve been able to obtain this information from the Nigerian Postal Service website and convert it to a form that can be used by other applications.
Imagine the possibilities – dating websites could use the postcode to identify the location of a person and match their profiles to people in their area or ecommerce websites that could use that to batch delivery of orders. The applications are enormous. Eventually we hope to be able to map center points for each of these postcodes even increasing the utility.
The data as it is, is exactly in the form it was extracted. We’ve identified a number of corrections to be made and we’re crowdsourcing (essentially getting interested fellows from the public) the cleanup of the data.
There’s a lot of data locked up on sites like the Nigerian Postal Service and what we’ve demonstated with the postcode data is that this data can be put to more use if it can be made open and made available in a form that computers can use.
You can access the data that has been put on Google Fusion Tables here.
July 24, 2010
8 responses to Nigerian postcodes – a case for open data
Hi Tim, i am very happy to find a great mind like yourslef interested in this subject. I wrote an article recently on my blog on how NIPOST and Nigerian Federal Survey can synergise to create an advance National Address database and also popularise the postcode system in Nigeria. In this article i proposed a NAG (National Address Gazetteer) that will underpin geographic decision made by Nigerian telecom, hospital, electricity provision, and government in all admistrative levels. Private companies in Nigeria like banks, insurance, etc can buy this data and attach their own data to it to evaluate the geographic spread of their operations. This can be a very lucrative business proposition for NIPOST and consortium and they could poetentially make enough money from it to make them stop going to Abuja with caps in hand every year soliciting for govt funds.____Read more about this article here: http://lagosstreetmap.blogspot.com/2010_09_01_arc…
This is remarkable stuff. Unreserved commendations must go out to you, Dipo & Kayode. I have been surfing the net for the last 3 hours in a frustrating roundabout fashion looking for this data. (I have also 'dreamt' of this data for about 2 years and had searched for it 'occasionally', while doing other stuff online).
There is so much that can & needs to be done with this data as you have rightly pointed out, especially in this buildup to the 2011 elections. As technology enthusiasts/professionals, we owe it to our country to contribute our quota to the salvation of this country (hope I'm not sounding too religious). It's the early hours of November 3rd 2010, and the mid-term elections in the US are rounding up. The media networks (CNN, Fox etc), the Political parties (democrats, republicans) are spitting out real-time mapped data as votes are collated in a country that is 10 times the size of Nigeria. Exit polls, historical voting patterns, election results etc. all laid out on maps. Please sign me up for the crowdsourcing intiative(s).
I got to know about Fusion Tables last week while doing a bit of research for Google DevFest 2010 in Lagos (http://code.google.com/events/devfests/2010/lagos.html) & I'm elated to discover that we have this sort of extremely valuable data already in the cloud. There's a lot we can do with this data, Google MapMaker, Maps API & Geolocation API, as well as utilities like OpenStreetMap (http://www.openstreetmap.org). I also discovered a 'paid for' (I haven't paid … yet) service from GeoPostcodes that offers this data inclusive of latitude & longitude for about 40 euros & the data looks pretty extensive (http://www.geopostcodes.com/sample_nigeria). You can actually browse through the data for free in a 'very little utility' way, but it costs 40 euros to gain access to the download site for 1 month. What do you think?
Once again, kudos for a job well done & an interesting blog post.
aniediudo [at] gmail [dot] com
Interesting comment Aniedi and thanks for the link to geopostcodes. I checked their data out but wasn\’t too impressed with their offering. Their geocoding is only accurate to the regional level which in my opinion isn\’t good enough for me to justify the investment but for the organization, that\’s great. I\’m thinking that we could use a service like Geonames (http://www.geonames.org/) to geocode a number of the locations in that database.
Hi Tim,
I am delighted about your interest and effort on geocoded address for Nigeria. I am also interested in your crowd sourcing, please sign me on. Meanwhile, most of the links are not are not responsive.
Hi tim , I know this has been a while since you were talking about this .. just wanted to bump this and say thank you for point me in the right direction … watch this space…
Hi Tim, did you notice the error in the postcodes published on NIPOST's website? (http://www.nigeriapostcodes.com/views/ ) e.g. I discovered that the postcode of Karonwi Street in Itire,within Surulere LGA is 100281. To my surprise, this postcode is shared by the popular Allen Avenue in Ikeja LGA!!
Could Surulere be that close to Ikeja to the point where they can actually share a common postcode? I created a map showing Karonwi street in Itire surulere and Allen Ave and measured the crow fly distance between the 2 streets using MapInfo Professional. The distance is 11.3km. Could a postcode parcel be that big and still remain relevant?
I am proposing a postcode using the following structure: (State + Postal Town + LGA + Electoral wards) see the rest of the article here http://lagosstreetmap.blogspot.com/2011/10/error-…
What do you think?
Pingback: Tekedia » The Error in Nigerian Postcode System – A Solution To Improve NIPOST Postcode System
Job well done