Session 14 — Data Driven
Saturday 09:30 - 11:00
Chair: Michael Pidd
This paper reports on the project entitled ‘Digitising experiences of migration: the development of interconnected letter collections’. It was funded by the AHRC’s research networking scheme from April 2013 to July 2014, and falls under their ‘Digital Transformations in the Arts and Humanities’ theme.
The project brings together scholars from different disciplines currently working with historical emigrant letters as a primary data source. It explores how letters are being used across the disciplines, identifying where there are similarities and differences in the transcription, digitisation and annotation practices being applied. Through developing a system of markup, specifically for emigrant letter collections, as well as exploring how digital technologies might be used to retrieve, search and visualise this markup, our aim is to initiate the process of interconnecting resources to encourage cross-disciplinary research.
In this paper we look specifically at the meta-information relating to author, recipient, location and date, within and across three large emigrant letter collections. First we will discuss how TEI (Text Encoding Initiative) markup was used to standardise this meta-information so that the collections could interconnect, and then we will demonstrate how visualisation tools were used to explore the meta-information in new and creative ways, for example by using GIS tools to map letter writing networks.
For further details about the project, please visit our blog: http://www.lettersofmigration.blogspot.com
University of Glasgow
Mapping Metaphor with the Historical Thesaurus is a linguistics and digital humanities collaboration which provides a data-driven picture of the extent and development of metaphor across the history of English. It uses a combination of database techniques and expert linguistic analysis to provide full details of the metaphorical links which speakers of English have used to understand the world around them for the past thousand years.
One of the project’s main outcomes is a database of tens of thousands of links between conceptual categories, such as imagination and biology (father, conceive, fertile) or beauty and light (shine, radiant, lustre), alongside date information and sample words. While there are a range of other outputs arising from this data, the online resource for exploring the links database – called the Metaphor Map – is a key output of the project. In addition to providing search and browse pathways through the data, there is also a need to provide a high-level visual overview for users. However, while such data is of huge relevance to scholars in a range of fields, including the study of language, literature, culture, and psychology, the size and complexity of the database provides a significant challenge with regard to displaying these connections to users in a way which makes their relevance clear.
This paper therefore discusses the means through which the Metaphor Map has been visualized by the team, including radial arc techniques, network diagrams, and treemap structures. We will demonstrate the importance of such visualisations to guide research on the database, and provide information from our feedback sessions on the understandability of such complex data. In so doing, we also aim to demonstrate the widespread, systematic and far-reaching impact of metaphor on English.
TransVis is a live online system for creating, curating, exploring and analysing corpora of retranslations -- i.e. differing versions of whole or part works e.g. by Aristotle, or the Bible, works by Shakespeare, etc. etc.. We work mainly with c.40 versions of Shakespeare's Othello in German (1766-2012). The system is language- and script-agnostic. Researchers with other corpora are testing the prototype at: www.delightedbeauty.org/vvv
Our ideal of digital access to retranslations combines public experiments in natural language computation, cross-cultural humanities debate, and playful user experience. We seek new ways to promote understanding of (inter)cultural diversity and change, past and present. We aim at enjoyable tools for small- and large-scale researches in world literature, scripture, philosophy, history of science etc. Also, to create interactive resources for teaching languages, translation, global cultural history and diversity -- and computational humanities techniques.
TransVis = Translation Visualization. Innovative visual interfaces enable users to:
1) create corpora (plain text or encoded)
2) define segment types, properties; segment texts; align segments
3) navigate parallel, aligned texts; search, filter, select, compare...
4) explore differences at levels from total structures to individual words (levels between: e.g. n-grams, defined segment properties)
5) explore intertextual relations via comparative algorithmic readings
Our prototype top-level interface, introducing a retranslations corpus, is an interactive time-map of sources, genres, connections. In future we want advanced metadata capture and mapping integrated with comparative analyses (e.g, stylometric comparison), plus curated text sample comparison. Currently we use GoogleTranslate; we would ideally crowd-source (re)(back)translations, commentaries, discussion. We have many desiderata, exciting possibilities.