Putting Medical Officer of Health Reports on the map - Natalie Pollecutt, Wellcome library

Editor's Notes

1. (Natalie) Debs and I plan to talk to you today about my rather low-rent method of mapping a set of our catalogue data using Google Fusion Tables. Google Fusion Tables is a Google tool that allows you to put your own data on Google Maps pretty quickly and easily without really having to know anything apart from your own data. I actually started on this when I heard about another tool for doing something similar, but after a bit of investigation found Google Fusion Tables. I already had a Google account so at least using this I didn’t have another username/password combination to remember. Caveats: Other tools are available
2. (Natalie) Start with a small data set e.g. people who have registered with our Library online, and then come into the Library to upgrade to full membership http://catalogue.wellcome.ac.uk:2082/screens/googlefusiontable.html
3. (Natalie) Where our online photography requests have come from, and been sent to http://www.google.com/fusiontables/DataSource?snapid=S270056IehV Having shown these around my department it was suggested that the Medical Officer of Health reports would be good candidates…
4. (Natalie) http://www.google.com/fusiontables/embedviz?viz=MAP&q=select+col1+from+494273+&h=false&lat=53.17311920264066&lng=-0.9633739406249386&z=6&t=1&l=col1 Click to show: Place Subject headings Link to catalogue, for info for ordering
5. (Natalie) And this is the Google Fusion Table behind it Table view: http://www.google.com/fusiontables/DataSource?dsrcid=494273
6. (Natalie) Click to show: Place Subject headings Link to catalogue, for info for ordering Many more records… But still a smallish and discrete data set Good geographical spread across the whole of the UK – which I didn’t actually know until I saw them on a map Debs – MOH reports
7. (Debs) -Key source for info on health of UK from mid 19 th – mid 20 th C Stem from reaction to infectious disease in mid 19thC MOH reports include information on such issues as mortality rates; disease outbreaks; sanitation; port hygiene; food inspection (ice creams & butter!); maternity and child welfare; housing; pollution; manufacturing (e.g. the inspection of workshops); shops and offices; demography; engineering  and meteorological conditions Vary from location to location: dependent upon interests of MoH officer for particular area Long regarded as an imp source for 19 th / 20 th history of Public Health Digitising these reports will improve access to important research material as well as offering opportunities for new approaches to text and data mining
8. (Natalie) I’d been to a training session (not about maps) where mapalist had happened to be mentioned. I’d looked at that at that’s what got me interested in the first place. When I started looking into it I wanted something that was: free web-based (wouldn’t have to install anything) easy-to-use, with support if available plus, I already had a Google account
9. (Natalie) I’ve now also made a few Google Maps with the version 3 API (very basic) with ‘Beginning Google Maps API 3’ by Gabriel Svennerberg http://www.doogal.co.uk/BatchGeocoding.php
10. (Debs) Our current MOH reports catalogue records are serial records and have limited information (impossible to search by year and geographical area) We are currently working on creating individual records for each report (as monographs) containing as much geographical information as possible (small parishes didn’t publish their own reports but statistics from them were included within a main authority report) This will enable researchers to search by geographical area, coverage dates, name of MOH as well as others We felt that being able to search by at least these three was important Plan to plot the timeline and maps One of the challenges is due to the administrative boundary changes of London being confusing Over the years the administrative boundaries of London changed and in 1889 the county of London formed from parts of the ancient counties of Middlesex, Surrey and Kent. This covered an area more or less what we now call Inner London In 1899 the County of London was split into 28 Metropolitan Boroughs In 1965 the current 32 Boroughs of Greater London were created The outcome we are aiming for is to provide a geographical history of an area showing boundary changes and areas of authority over time (so current borough records will include the place names for any preceding parishes or metropolitan boroughs)
11. (Natalie) 1) Some way to easily and programmatically geocode a lot of data, preferably from the sorts of place names that our library catalogue records hold, LCSH?? 2) What to do when we have one record but multiple place names – how do we handle those situations?
12. The other problem is that even when you do get the correct location, e.g. London, Google Maps just layers all the markers for all the records on top of each other, so you only see the marker for the latest record in the table and none of the ones before it. I went in an hand-changed a lot of markers (mainly by finding nearby postcodes). What’s right for catalogue records (consistency in place names via authority files) isn’t very helpful when you’re trying to put multiple markers for the same place on a map, or when boundaries change over time. Has anyone addressed this with catalogue/database data? Has anyone got any good ideas? Google suggest adjusting the longitude and/or latitude, which can be done, but I’m not sure it’s the best approach I did learn some stuff about mapping our data, particularly what issues we would face if doing the same again.