Katrina Navickas

Winner's Blog Posts:

Submitted Entry for 2015 Competition

Abstract


Protest and popular politics is about space and place. For example, Occupy is just the latest among a long history of social movements who make a political point through contesting the uses and meanings of public spaces. But for historians and sociologists of social movements, our main evidence of protests is textual, not spatial. Historians have so far only been able to plot the locations of small numbers of historical political meetings manually. Samples and scale are by necessity small and thus unrepresentative. We know that urban development and changing street plans shaped the locations and forms of political meetings in the past, but it is much more useful to visualise these changes and patterns geographically on a large scale.

What if we could have a tool to extract notices of meetings from the British Library collection of historical newspapers and automatically plot them on layers of historical maps? We could visualise the locations of political events using a significant sample than is manually possible. We could see new spatial patterns in where protests happened, and in so doing, help answer the question why they happened. Regions, towns, even streets could find a longer sense of their political heritage, enabling them to find out what meetings or events occurred in their area, and therefore encourage a continued engagement with politics and democracy among local communities.

This project showcases three of the British Library's collections and combines them in a way not done before: the geo-referenced maps in the BL geo-referencer and Flickr commons collections, the Ordnance Surveyors' drawings, and the 19th century newspaper collection. It will develop a tool for text-mining and geo-locating the records of political meetings, initially from the 1830s and 1840s from the Northern Star newspaper, and plotting them on layers of historic maps. The public will access the maps and data, and the tool will be adaptable to enable scholars to plot any form of event and spatial information using historical texts and maps.

URL for project:
http://protesthistory.org.uk/?page_id=485 for a diagram of the process and for how I already use similar data and maps.

Assessment Criteria

The research question / problem you are trying to answer

Please focus on the clarity and quality of the research question / problem posed:

- How can historians and scholars of social movements find and interpret geographical concentrations and clusters of political meetings and protests recorded in historical newspapers?

- In finding these geographic patterns, what do they tell us about levels of political activity in different areas? Where were democratic movements like the Chartists meeting? What was the average size of their meetings, and how were they constrained by their urban environment? Which sites have continued their use as sites of public meetings to the present day?

These are the central questions that this project seeks to answer using a specific case study of political meetings recorded in the Northern Star newspaper (1838-44) and regional and town maps of Britain in the British Library collections dating from the 1830s and 1840s, with the goal of expanding to a wider range of newspapers and maps.

Geographers and urban planners have long used digital methods in their research. The Centre for Advanced Spatial Analysis (CASA) and the Space Syntax Laboratory at the Bartlett School of Architecture, University College London (http://www.bartlett.ucl.ac.uk/casa ), for example, are developing pioneering ways of understanding the spaces of modern cities. This project seeks to show that historians aren’t being left behind in these developments. Historians are beginning to take ‘the spatial turn’ with digital tools. The Locating London's Past project (http://locatinglondon.org/ ) has shown the potential of plotting crime and poor law records from various sources on an 18th century map of London. Lancaster University's Geography department is also examining the possibilities of combining corpus linguistics with GIS in their Mapping the Lakes project to map romantic travel writing from the early 19th century (http://www.lancaster.ac.uk/mappingthelakes/ ). Radical Meetings Mapper seeks to emulate the spirit of enquiry and digital techniques of Locating London's Past, using a different type of source - local newspaper records - and different subject matter - protest meetings in the early 19th century, the start of the first campaigns for democracy in Britain. It seeks to find confirmation of expected patterns (e.g. that certain pubs in industrial areas were always used for Chartist meetings; that the more outlying 'neighbourhoods' and rural villages around towns were most active in political activity campaigning for the vote).

But more importantly, the project also aims to use the visualisations and natural language processing to generate new questions as well as answers that will not simply benefit this project, but lead to new research in many other fields that could benefit from a tool to extract and plot textual data onto historical maps.

Please explain the ways your idea will showcase British Library digital collections

Please ensure you include details of British Library digital collections you are showcasing (you may use several collections if you wish), a sample can be found at http://labs.bl.uk/Digital+Collections

The tool will draw its historical data from the 19th century newspaper collection, but its main aim is to showcase (and indeed add new uses and interactivity to) the rich collection of digitised maps, combining the geo-referenced map collection (http://www.bl.uk/maps/georeferencingmap.html ) with the maps tagged in the images collection on the BL Flickr commons account and the Ordnance Surveyors' Drawings of England and Wales (1789-ca.1840).

Connecting these several collections is a major aim of this tool, and thus distinguishes it from previous winning entries of the BL Labs competition, which have mainly focused on showcasing only one BL collection.

The project will visualise the results in a web interface of layered historical maps, on which the extracted data of political meetings is plotted. It would be great if the British Library could host the interactive layers on their Geo-Referencer pages of their website, but if this is not possible, I will host it on my own website (http://protesthistory.org.uk ) which I hope to migrate to the new website of the University of Hertfordshire Centre for Digital History Research to ensure sustainability. The public will engage with the maps by turning off and on the different map layers to see change over time, and filtering the plotted meetings by type and date. Clicking on the data points will reveal the data (place, type of meeting, date, source and link to the original digital content).

Please detail the approach(es) / method(s) you are going to use to implement your idea, detailing clearly the research methods / techniques / processes involved

Indicate and describe any research methods / processes / techniques and approaches you are going to use, e.g. text mining, visualisations, statistical analysis etc.

(see diagram at http://protesthistory.org.uk/?page_id=485 )
stage one - text mining the BL 19th century newspapers collection.

The first sample will be taken from the OCR of the Chartist newspaper the Northern Star , 1838-44, which has regular and standardised columns of 'forthcoming meetings' which can be identified automatically through natural language processing. There are about 15 to 20 meetings listed in each weekly paper (more if it is possible to identify meeting sites from other columns). This will produce a sample of c.5000 records for places across Britain, with the following data: date, place, address, type of meeting (democratic, Chartist, socialist, trade union, anti-radical, riot), time, and source.

See the diagram and sample at http://protesthistory.org.uk/?page_id=485
If the OCR proves too problematic , or if we are unable to access the xml underpinning the digital database, after the sample has been selected, I will use the Omeka-Scripto transcription platform to enable my existing team of volunteers to transcribe the basic information from the texts with xml-tags to identify the meeting sites and types of meeting.

stage two - selection of geo-referenced maps.

The main maps of British regions and towns from the 1830s and 1840s will be selected from the BL geo-referenced map collection, and also the Ordnance Surveyors' Drawings of England and Wales.

stage three - building a gazetteer using current geo-coder (e.g. http://www.findlatitudeandlongitude.com/batch-geocode/ ) and historical gazetteers from BL collections and other sources (e.g http://www.gazetteer.org.uk/ , http://www.placenames.org.uk/ )

stage four - developing the tool to geo-locate and plot the meetings in the sample using the gazetteer and the geo-referenced map layers

stage five - developing the web interface to visualise the plotted points on the layers of historical maps and making the maps appealing and useable for the public.

Please provide evidence of how you / your team have the skills, knowledge and expertise to successfully carry out the project by working with the Labs team

E.g. work you may have done, publications, a list with dates and links (if you have them)

I am senior lecturer in history, with a research specialism in popular protest in eighteenth and nineteenth-century England. I am a member of the new Centre for Digital history research at the University of Hertfordshire.

Plotting and looking for spatial patterns in protest sites is at the heart of my research. My forthcoming second monograph, Protest and the Politics of Space and Place, 1789-1848 will be published by Manchester University Press later this year. I am trained as a historian in more traditional ways, so, seeing the opportunities that digital humanities now offers, I have been teaching myself how to use digital tools and methods over the past three years. To augment the standard historical research for my book, I have built up an online database of sites at: http://goo.gl/2F5drh . I visualised some of this data on my website using google fusion tables, maptiler and other open source software. See for example the maps of Manchester on http://protesthistory.org.uk/?page_id=25

My technical skills are self-taught, including GIS (I mainly use QGIS for geo-referencing and drawing maps and plotting my data, though I have taken a short course in ArcGIS, and I have played around with different platforms including CartoDB), and basic competency in xml and html.

I am also very keen on encouraging public access to and engagement with historical documents. I was the CI in a British Academy-funded pilot project with the National Archives in 2013-14, in which we catalogued and digitised 25 boxes of Home Office documents from 1816-17 (c.25,000 images) and are now in the process of getting them transcribed and xml-coded, with the aim of integrating the database with the other collections on British History Online. I set up a crowd-sourcing transcription website using the Omeka platform and the Scripto plugin - http://transcribethehomeofficepapers.net/ to develop the next stage of the process. The project enabled us to build up a group of committed volunteers, and I have also employed undergraduate students in transcriptions as part of their work experience module, so I am able to draw from a team if the OCR'd and marked up text needs to be checked and corrected.

I am by no means a 'digital historian' of multiple competencies yet, but I hope to be in the future. I understand the main language of the technologies involved, and moreover would relish this opportunity to work with your team to develop my technical skills further.

Please provide evidence of how you think your idea is achievable on a technical, curatorial and legal basis

Indicate the technical, curatorial and legal aspects of the idea (you may want to check with Labs team before submitting your idea first).

Technical
All the main stages of the process are technically feasible and should not require a large amount of assistance. The methods for each stage exist already, and the challenge for the BL Labs team will be to develop the tool to connect the different stages together. The sample of one newspaper for 1838-44, with clear columns in generally the same place in each issue will make the identification of the data a manageable task within the timeframe, though I hope to expand the tool to incorporate different newspapers and data to make it useable for other scholars interested in other research topics. The mapping side of the project is even more simple, especially as we will be using already geo-referenced maps and the BL's web interface, although there will be more development needed for the tool to plot and filter data. I also hope to seek advice from the technical team for Locating London's Past, who are experienced in mapping historical data.

Curatorial
I am solely using British Library collections and metadata, accessible on site at the British Library. I will need access to the OCR'd text and xml mark up of the 19th century newspaper collections, so I will need to liaise with the BL newspaper curators.

Legal
The historical maps are out of copyright and under the British Library's control. I understand that access to the the 19th century newspaper collections is not as straight-forward, but I won’t be posting the original images, so there should be no legal obstacles, though the BL Labs team will consult with Gale Cengage to clarify use of the OCR’d and marked-up text.

Please provide a brief plan of how you will implement your project idea by working with the Labs team

You will be given the opportunity to work on your winning project idea between June 2015 - October 2015.

June 2015
a) Planning with the BL Labs team and the curators of the BL digital newspaper archives and maps collection as to the scope and processes involved in the project, including the technical challenges and processes, and sustainability beyond the development time.
b) BL Labs team to consult with Gale Cengage about drawing data from the 19th Century Newspapers Database.
c) Initial selection of geo-referenced maps in consultation with the BL Labs team and map curators.

July 2015
a) Working with the BL Labs technical lead to develop a natural language processing tool to text-mine the sample of the Northern Star newspaper in the 19th Century Newspapers database.
If there are difficulties with automatically text-mining the newspaper text, I will work independently with my existing team of volunteers from my Home Office project to transcribe and mark up a sample of newspaper reports.

August 2015 (main phase of development)
a) By me and if necessary, my volunteer team, continued marking up in xml of the newspaper reports.
b) Layering of the historic maps and development of the web interface for displaying the maps, and building a gazetteer using current geo-coder with the BL Labs technical lead.
c) With the BL Labs technical lead, developing the tool to geo-locate and plot the meetings in the sample using the gazetteer and the geo-referenced map layers.

September 2015
Independent work on developing the web interface to visualise the plotted points on the layers of historical maps.

October 2015
a) Testing the web interface with my team of volunteers.
b) Consulting with the wider BL community to promote the map data and the mapping tool.
c) Working independently to monitor the usage of the maps, promoting the maps via social media and my own website