CfP #dhiha7: Text as a resource. Text Mining in Historical Science
Conference june 29th and 30th 2017.
“What do you do with a million books?” – Gregory Crane’s question does not only raise questions about the conditions in which the Humanities acquire knowledge. It can also be seen as indicating a methodological turning-point, unfolding today, coming from the discipline of Ancient History. It consists of all the efforts, trying to make the increasing amount of digital data, structured or unstructured, available for research in the Humanities. Making available in this sense has two sides: one is about making data accessible for scientists (e.g. digitization, retro-digitization), the other is about providing tools which allow to dig into the data, especially into huge amounts of data, for analytical purpose (Text Mining, Data Mining).
So far, Text Mining has mainly been used in philology, in order to – for example – attribute texts of a certain phrasing or word frequencies to a specific author, or to characterize typical language patterns of certain genres, eras or works. Being confronted with semi-automatic analysis, Historians, referring to the hermeneutical tradition of their subject, usually remain rather skeptical. This attitude might change, the more data for historical research is available and the more projects with a specific outcome have been accomplished, showing the value of Text Mining when it comes to answer certain research questions.
In order to unfold the potential of Text Mining within historical research, the availability of appropriate tools, providing and integrating frequency analysis, co-occurrence analysis, topic modelling, sentiment analysis, annotation procedures, machine learning and others, is crucial. Until today, research funding has already allowed to develop these tools and infrastructures, which exist and which are working. A great part of them also allows the use of Text Mining-tools without necessarily requiring a deeper or any understanding of programming or computer sciences. This is important, because our conference is mainly based on the assumption, that requiring competences in programming or computer sciences would create an additional barrier for Historians trying to use digital methods or could even prevent them from doing so.
Therefore, the conference has two major aims. One is about presenting existing infrastructures, tools, applications and analysis strategies for the use of Text Mining in specific research contexts. The other is about presenting research projects having applied Text Mining, which will allow us to show how to apply these methods and what outcome this could mean. These two aims taken together, the conference will allow for researches in Historical Science to get an overview of existing techniques and strategies for mining huge amounts of text data. As infrastructures, tools and projects – given the predominant logic of national research funding – are most likely not know beyond their national scientific system, we also aim at bringing together a wide-range, international group of scientists in this conference, allowing exchange and cooperation.
Therefore, we are especially interested in contributions, which address the following fields of research:
- Research infrastructures for Text Mining (what functions are available, what is their current status, regarding development and accessibility, etc.)
- Strategies for an appropriate application of Text Mining (consecutive steps or parts of analysis, data processing, blended reading, etc.)
- Presentation of research projects in any field of History, having used Text Mining in various forms (e.g. sentiment-analysis or others)
- Legal aspects (copyright, access to data and corpora, re-use of data and corpora, etc.)
- Reflections on epistemological changes in Historical Science due to the application of Text Mining
This conference will be the seventh in the series of “Digital Humanities at the GHI”, which received great international attention and which has been presented by the Humanities Department of GHI Paris from 2011 on. The conferences have been and are organized with various partner and funding institutions. They also get a live coverage in a German and French speaking science blog. For this seventh conference, the GHI has the honor to cooperate with INRIA, the French Institut National de la Recherche dédié au Numérique, and its research team ALMAnaCH, which is specialized in the automated analysis of language.
Paper submissions (in English) –300 words max. – can be sent to à email@example.com until February 28th 2017.