Session 4

Friday 09:30 - 11:00

High Tor 3

Chair: Jamie McLaughlin

Industrial Memories: Data Analytics as a form of Witnessing

  • Emilie Pine,
  • Mark Keane,
  • Susan Leavy

University College Dublin

This paper will address the ethics of using human-centric digital methodologies to intervene in cultural memory as an act of belated witnessing.  

 

The Irish Government’s Commission to Inquire into Child Abuse published its final report in 2009 (henceforth, the Ryan Report). The Ryan Report detailed the history of child abuse across 50+ institutions in the Irish State, from 1940-1999; the 2,600 page report gathered evidence from 10 years of inquiry, involving over 1500 witnesses. Though the report contains key facts about how institutional abuse can arise and the failure of governmental oversight, its sheer size has meant that the totality of its findings have never been properly communicated to the public.

The Industrial Memories project treats the Ryan Report as a data corpus which can be mined and analysed – the panel will particularly focus on the intellectual/ethical principles of treating this material as data, the emotive impact of the material, the methodology of approaching a governmental report as a corpus, the generation of research questions, and the potential outcomes of identifying behavioural, cultural and linguistic patterns.

This paper will give an overview of the Industrial Memories project and its mission to digitally re-read and re-present the Ryan Report. Pine will discuss the 2 different strands of the project: 1) the text and data analytics approach, and 2) the digital arts strand: the mobile app walking tour, virtual reality platform, and GIS/heat map. Pine will discuss ways in which the data analytics project enables the scaling up of traditional humanities questions, the challenges to this (the risk of flattening data), and the combination of data approaches with memory and trauma theory. As a performance and literary scholar, Pine will examine the promise of ‘digital reading’ and how digitally re-reading enacts a different kind of memory work. In the second half of the paper, Pine will present an overview of results so far, including an introduction to ‘Transfer Analysis’ and ‘Witnessing Analysis’.

The Industrial Memories project is funded by an Arts/STEM major grant from the Irish Research Council, 2015-2018.

Dr Emilie Pine lectures in modern drama at University College Dublin. She is the founding Director of the Irish Memory Studies Network, PI on Industrial Memories (IRC New Horizons 2015-18), Assistant Editor of the Irish University Review, and author of The Politics of Irish Memory (Palgrave, 2011). Emilie is currently working on a book on memory, the witness and the marketplace in world theatre.

Industrial Memories: Methodologies for analysing the Data of the Commission to Inquire into Child Abuse Report (Ryan Report)

  • Emilie Pine,
  • Mark Keane,
  • Susan Leavy

University College Dublin

The Irish Government’s Commission to Inquire into Child Abuse published its final report in 2009 (henceforth, the Ryan Report). The Ryan Report detailed the history of child abuse across 50+ institutions in the Irish State, from 1940-1999; the 2,600 page report gathered evidence from 10 years of inquiry, involving over 1500 witnesses. Though the report contains key facts about how institutional abuse can arise and the failure of governmental oversight, its sheer size has meant that the totality of its findings have never been properly communicated to the public.

The Industrial Memories project treats the Ryan Report as a data corpus which can be mined and analysed – the panel will particularly focus on the intellectual/ethical principles of treating this material as data, the emotive impact of the material, the methodology of approaching a governmental report as a corpus, the generation of research questions, and the potential outcomes of identifying behavioural, cultural and linguistic patterns.

This paper will discuss the careful design of a database to capture aspects of the reports contents; a design that lends additional structure to the current text of the report, without over interpreting its contents for those who wish to explore it further. Leavy will address the development of a web-based exploratory interface, enabling both close and distant readings of the Ryan Report. Text mining tools such as concordance analysis, detailed text searches, statistical analysis of the text and keyword analysis will be demonstrated to illustrate how these were used to uncover patterns and insights in the text. The creation of an abstract representation of knowledge contained in the report through the searchable web interface will be described, demonstrating how each abstraction also links back to occurrences within the text to facilitate exploration of text extracts in context. The paper will conclude by outlining initial findings of the current research relating to passive and active language and how this exploratory interface might be applicable to other research data/fields.

 

Dr Susan Leavy is a postdoctoral research fellow on the Industrial Memories project, funded by the Irish Research Council 2016-18 at University College Dublin. Her PhD research analysed gender bias in media coverage of politicians in Ireland using natural language processing and machine learning. She has an MPhil in Gender and Women’s Studies from TCD, an MSc in Artificial Intelligence from Edinburgh University and a BA (English & Philosophy) from UCD. She also worked internationally for several years in technology in the finance industry, on the development of global trading platforms. She is currently working on Industrial Memories, a project using digital methods to analyse the findings of the 2009 report on abuse in Ireland’s industrial schools (the Ryan Report).  

Co-author: Professor Mark Keane

Dr Mark Keane is the Chair of Computer Science at University College Dublin (since 1998). From 2004-2007 he was Director of ICT (2004-2006) and Director General (2006-2007) at Science Foundation Ireland (SFI) and was also VP of Innovation & Partnerships at UCD (2007-2009). He has a BA (UCD) and PhD (TCD) in Cognitive Psychology and previously worked in University of London, the Open University, Cardiff University and Trinity College Dublin (FTCD, 1994).  He is co-author with MW Eysenck of “Cognitive Psychology: A Student’s Handbook” which has been translated into 5 languages and is now in its 7th edition.  In recent years, he has worked as a PI in Insight on text analytics problems; including tools for digital journalism, the use of news to predict market shifts and bubble behaviour and an analysis of the problem of farm land succession in Ireland.


The Industrial Memories project is funded by an Arts/STEM major grant from the Irish Research Council, 2015-2018.

Worldbuilder: a tool for Text-World annotation and visualization

  • Yufang Ho,
  • Dan McIntyre,
  • Jane Lugea,
  • Jing Wang,
  • Zhijie Xu

University of Huddersfield

In this presentation, we demonstrate Worldbuilder, a tool that we have developed at University of Huddersfield, UK, for text-world analysis and visualisation. Worldbuilder allows the user to annotate uploaded texts using an annotation scheme that encompasses the range of descriptive categories found in Text World Theory (Werth 1999; Gavins 2007). The annotation scheme incorporates the range of sub-categories associated with world-building elements, function advancing propositions and world switch possibilities. Once a text has been annotated, Worldbuilder can be used to produce visualisations of the annotated data in the form of text-world diagrams. To demonstrate the software, we present a brief analysis of data from the case of Amanda Knox, whose murder trial in Italy attracted worldwide media attention from 2007 to 2015. We use Worldbuilder to annotate and analyse the text-worlds projected in the controversial statements in the case under study, and then compare the cognitive structures of the statements at the macro level and examine their individual linguistic features at the micro level. We suggest that Worldbuilder enables the systematic analysis and visualisation of longer stretches of texts than is possible manually, and that this expands the range of texts it is possible to analyse and the purpose for which such analyses are carried out.

 

Gavins, J. 2007. Text World Theory: An Introduction. Edinburgh University Press.

Werth, P. 1999. Text Worlds: Representing Conceptual Space in Discourse. Longman.