NCRM International Visitor Exchange Scheme (IVES)

Machine Reading Newspapers: Natural Language Processing Methods to Build a Global Labour Conflict Database (2016 - 2017)

Dr Juan Grigera (Institute of the Americas, University College London) (j.grigera@ucl.ac.uk) visited Prof Beverly Silver (Department of Sociology, John Hopkins University).

Up until recently the possibilities of studying social phenomena from a global and historical perspective were limited, and thus its importance was only partially acknowledged. The rapid change in the availability of data (or the growth of so-called Big Data) is looming large in the minds of social scientists, as well as the even more apparent limits of national-centred approaches. In the case of social unrest, how to gather and assess empirical data has been for a long time a key methodological issue. The limited geographical and temporal scope of most time series and their incommensurability, in particular of those few governmental-collected available sources. This visit to the well-established Global Social Protest Research Working Group at the Arrighi Center for Global Studies at John Hopkins University in Baltimore (US) is set to further advance and intertwine the recent methodological breakthroughs stemming both the disciplinary domains of global social history and natural language processing.

Main objectives of the visit:

• to develop an automated event reader that extracts structured information suitable for analysis from newspaper sources and is integrated with feedback to an ongoing workflow of data collection

• to become acquainted with the multiple complexities of newspaper data collection and analysis for the uses of global social history

• to learn about the possibilities of using NLP methods by intensively applying them to a specific domain