We give notice of the on-line workshop Sentiment Analysis in Literary studies organized by the Centre for Information Modelling of the University of Graz.
Sentiment analysis is a common task in literary studies, yet sitting outside the mainstream of analytic computational procedures applied to philosophical corpora. Critic facets of sentiment analysis procedures for historical-philosophical analysis lie primarily on tools’ dictionary-dependancy, from which follow difficulties in obtaining in-depth historical understanding and the possibility of arbitrary biases in interpretation of both the retrieved sentiment and its object. However, reasons for such a sidelining hold when they are referred to techniques and workflows commonly deployed and followed in order to achieve sentiment analysis, while wanting to perform such a task might not be a radically flawed endeavour per se, as long as researchers set a well-grounded research framework.
The workshop’s programme anticipates a thorough examination of existing tools, approaches and workflows as well as of preliminary steps such as textual preprocessing, and it is uncommonly devoted to the analysis of 18th century texts.
The workshop fills in the context of the project “DiSpecs – Distant Spectators. Distant Reading for Periodicals of the Enlightenment”, funded by CLARIAH-AT and in cooperation with the Institute for Interactive Systems and Data Science, the Know-Center GmbH Graz and the Centre for Information Modelling – Austrian Centre for Digital Humanities (ZIM-ACDH) and the Institute for Romance Studies. The project aims to investigate the digitized, TEI encoded and semantically enriched texts of The Spectators (http://gams.uni-graz.at/mws), with quantitative methods of data analysis referred to as distant reading and macroanalysis (topic modeling, stylometry, meme diffusion, sentiment analysis and community detection).
The 18th-century journalistic genre of “spectators” (or moralistic sheets) had a large audience of urban readers and played an essential role in public opinion genesis. This project endeavours to create an integral database for all the moralistic press in French, Italian, Spanish, English, German, and Portuguese. In this context, the Spectator discourses’ quantitative analysis aims to enhance and improve micro-narration studies regarding the repetition of motifs throughout different journals.
Official information is reported below.
Sentiment Analysis in Literary Studies, 2021
Online, February 17-19, 2021
The workshop introduces the concepts of Sentiment Analysis and will give an overview of related methods and tools with a special focus on their application to historical literary text corpora. The participants will be presented with the following content:
- Introduction to sentiment analysis: methods, projects, tools, and first steps
- Hypothesis testing
- Challenges of sentiment analysis in historical literary texts
- Dictionary-based sentiment analysis
- Preprocessing steps
- Using a tool chain
- Project presentations of the participants
Each workshop day will feature a keynote lecture from experts in the field. All keynotes will be free of charge and open to the public.
Participation in the hands-on workshop is free of charge and open to 20 students and scholars of all academic stages. No previous specific skills are required (although general computer literacy is expected).
To apply, we ask you to use the application form and provide a brief motivational note why you would like to attend the workshop (max. 250 words).
We reserve the right to choose the individual participants based on their research/study interests and motivation.
Applications are welcome until January 15, 23:59 CET.
Acceptance will be communicated by January 25.
Bernhard Geiger (Know-Center Graz)
Christina Glatz (University of Graz)
Elisabeth Hobisch (University of Graz)
Philipp Koncar (Graz University of Technology)
Sanja Sarić (University of Graz)
Martina Scholger (University of Graz)