Explorations in Textual Digital Analysis for the Humanities and Social Sciences
5th Workshop of the Research Programme “Clash of Civilizations or Peaceful Co-evolution? Intercultural Contact in the Age of Globalization”
14-15 May 2015
The Institute for Global Studies, in collaboration with DigiHUBB – The “Transilvania” Centre for Digital Humanities, Babes-Bolyai University, Romania, invites national and international scholars to convene for 2 days, to discuss, explore and develop their technological and humanistic imagination in relation to text analysis.
Given the continuous online content development, archiving, storing and presentation combined with the relatively recent efforts of public and private institutions to digitalize public domain texts and documents from different historical times, digital textual analysis is an increasingly productive research method in the social sciences. In this workshop, we address the principles, methods, and problems of exploring historical texts as data, with particular relevance to intercultural conflict and harmony in the historical regions of Banat and Transylvania.
Text analysis is a broadly defined research method to reveal semantic structures and relationships, textual patterns and statistics. In the late 1950s, H. P. Luhn published a paper detailing his “exploratory” research based on excerpts of technical papers and magazines with the ultimate goal of producing abstracts. His research consisted of text mining and analytics in which the “machine” would measure the significance of words and sentences based on analytical information of frequency and distribution. Some 60 years later, a multitude of computer software stands ready to help the scholar with text analysis.
Text analysis enables the investigation of vast amounts of text, text as data. Our workshop will explore the most productive methods of digital textual analysis currently available and applicable in cultural historical research, such as:
Text Mining / Data Retrieval. OCR (optical character recognition) is a method for extracting the text from a digital document. Methods of text mining are available through different software or through the creation of software. A significant part of text mining software has been developed by public research and educational institutions and are available online in open source. The Metadata Offer New Knowledge (MONK) has been developed at the University of Illinois; Overview is a tool for organizing documents and text files in different formats and visualize patterns of connections based on word frequencies and topic tagging; Google Ngram Viewer offers the possibility to search terms and concepts within the works available online with Google Books; the Natural Language Processor and Analyzer (NLPA) was developed at the University of Indiana to discover text analytics.
R and Digital Text Analysis. The programming language R has been built with the primary purpose of statistics. However, it has also been frequently employed in research methods for digital text analysis in the social sciences. R enables text analysis at all levels, from analytics to pattern recognition and visualization.
Sentiment Analysis. Sentiment analysis offers the possibility to identify, catalogue and classify attitudes discovered within texts towards a given issue within a simple or complex set of classes with available training data for supervised learning (i.e. good, negative) or unsupervised learning based on semantic orientation of phrases available within documents. Sentiment analysis may thus be profitably applied in the study of intercultural harmony or conflict.
Network Analysis and Information Visualization. Information visualization enables us to visually observe interactions of data in vast amount of data sets that would otherwise be missed. Network Analysis allows us to understand the level of influence of actors and the distribution of concepts within data sets. NodeXL or Gephi are viable tools for network and information visualization
The workshop, in addition to serving as a platform for scholars who conduct research employing some of the foregoing methods, will also explore recent contributions to the development of textual analysis software and its possible applications in the humanities and the social sciences.
Contributions and participants
The workshop is open equally to users and to developers of textual analysis software.
We invite contributions of theoretical essays, discussion papers and empirical studies or demonstrations of the various uses of textual analysis software.
Please submit a title and abstract (300-500 words) by 15 April.
Full papers of up to 6000 words are expected by April 30, 2015 to fofiu at fspac.ro. Revised papers will be published in an international volume by the end of 2015.
Venue and programme
The workshop will take please in room T2 at the Faculty of Political, Administrative and Communication Sciences, Babes-Bolyai University (85 Minerilor street, Cluj-Napoca).
The programme of the event is available here
Prof. Mihai I. Spariosu, University of Babeș-Bolyai, University of Georgia, Athens
Prof. Gheorghe Stefan, Polytechnic University of Bucharest
Dr. Adela Fofiu, Babes-Bolyai University, Cluj-Napoca
Dr. Corina Moldovan, DigiHUBB, Cluj-Napoca