This two-day symposium aims to bring together scholars and researchers working with computational approaches to texts. The event targets a broad audience interested in the application of digital text analysis technology (text mining, topic modeling, authorship detection, tasks performed through Natural Language Processing) for Resource-Scarce Literatures and from an historical perspective. Come join us!

PROGRAM

Wednesday, April 24

9:00-9:30AM

Coffee & Breakfast

9:30AM-12:30PM

Workshop 1
Modelling Hogwarts. A speed introduction to digital text analysis.
Taught by Mike Kestemont (University of Antwero) & Enrique Manjavacas (University of Antwerp)

In this three-hour workshop, we offer a speed introduction to a number of established approaches in digital literary studies. The case studies are taken from the universe of Harry Potter but the methodologies are generic enough to be applicable across different domains of the Humanities. We will focus on the following two applications:
1. Network analysis: Fanfiction is intriguing because of the new combinations of characters and themes it introduces. We will use network analysis (Gephi) to model these remarkable (and often radical) shifts in attention with respect to the original canons. This macro-level approach can be considered a modern form of reader response analysis.

2. Stylometry: Authorship attribution is a popular research topic in stylometry, the quantitative study of writing style. We will use "Stylometry with R" to replicate a famous case study: J.K. Rowling secretly published a novel under the male pseudonym Robert Galbraith, but Patrick Juola unmasked the author using stylometry. We will analyze how genre and literary influence are key factors in modelling an author's writing style.

The workshop will be highly hands-on and people are expected to bring their laptops on which they can install new software. Installation instructions for the software will be sent out to the participants in the weeks before the event. All data will be distributed on-site.

12:30-1:30PM

Lunch

2:00PM-5:00PM

Workshop 2
Lexos for Analysing Historical Literatures in the Classroom.
Taught by Scott Kleinman (California State University, Northridge) & Mark LeBlanc (Wheaton College, MA)

This workshop will demonstrate the capabilities of Lexos for analyzing literatures in historical languages. Lexos is a web-based tool designed for transforming, analyzing, and visualizing texts. Lexos was created by the Lexomics Research Group (Michael Drout at Wheaton College, Scott Kleinman at Cal State Northridge, and Mark LeBlanc at Wheaton College) as an entry-level platform for Humanities scholars and students new to computational techniques while providing tools and techniques sophisticated enough for advanced research. It has been designed specifically to include rich pre-processing features for use in the study of historical languages. Lexos is an open-source web-based tool written in Python, which been under development for ten years with the support of three NEH grants and a unique model in which undergraduates serve as both researchers and software developers. The project has resulted in numerous faculty and student-authored publications (listed here). Lexos may be installed locally; however, we also maintain a live, server-based installation accessible over the internet, making it ideal for classroom use. The tool is accompanied by an embedded Scalar Book (“In the Margins”), which is designed to provide information on algorithms and best practices in computational text analysis. Although “In the Margins” is still under development, we continuously add content (and welcome external contributions) to it, which will ultimately enable users of Lexos to find relevant discussion about the analysis of historical languages embedded within the tool’s workflow. We propose a 2-3 hour hands-on workshop, in which we will provide an overview of the available features in Lexos, with special attention to functions related to handling premodern texts (e.g. Unicode, XML and HTML entity conversion, TEI markup handling, form consolidation, and lemmatization). We will demonstrate the use of Lexos with some case studies using Old English, Middle English, and/or Latin literature. The case studies will be based on successful experiments used in the classroom or in student research projects to showcase the accessibility of Lexos for teaching early literature as well as computational stylistics. Workshop attendees will be provided with the opportunity to try out their own data sources and to discuss how the tool could be improved to better support the study of historical texts.

Location
Richter Library 3rd Floor Conference Room
1300 Memorial Dr
Coral Gables, FL 33146)

Thursday, April 25

9:00-9:30AM

Registration & Breakfast

9:30AM-12:45PM

Talks Session 1

Mike Kestemont (University of Antwero), The case of fanfiction in Digital Humanities research and pedagogy -- Because the story's not over until they say it is.

Gimena del Río Riande (IIBICRIT-CONICET, Argentina), Computer-assisted Semantic Annotation not just for DH scholars

(break)

Greta Franzini & Marco Passarotti (Università Cattolica del Sacro Cuore) and LiLa (Linking Latin): Piecing together a seemingly under-resourced language)

Dennis Tenen (Columbia University), Minimal Computing in Plain Text

1:00-2:00PM

Lunch

2:00PM-5:15PM

Talk Session 2

Francisco Gago Jover (College of the Holy Cross), The Old Spanish Textual Archive, or what to do when you find yourself with 30 million words

James Gawley (University at Buffalo, SUNY), Unsupervised Lemmatization Model for Underserved Languages

(break)

Scott Kleinman (California State University, Northridge), Some failed attempts at lemmatising Middle English: what went wrong and where do we go from here?

Enrique Manjavacas (University of Antwerp), Lemmatization of Historical Languages with Neural Networks

5:15PM-5:45PM

Closing Discussion

This symposium is sponsored by the Univerisity of Miami College of Arts & Science, Department of Modern Languages & Literatures Joseph Carter Memorial Fund, SEEDS, School of Communications, Center for the Humanities & Digital Humanities Flanders.

Computer-Assisted Text Analysis For Resource-Scarce Literatures Symposium