Dieser Webauftritt wurde archiviert und wird nicht mehr aktualisiert.

Bei technischen Fragen kontaktieren Sie bitte das Webteam. Es gilt die Datenschutzerklärung der TU Darmstadt.

Text Analysis and Mining for Digital Humanities


In recent years there has been a growing effort to digitise humanities and cultural heritage data. Many large-scale digitisation projects are either underway or have already been completed. Museums and archives are making more and more of their data accessible online. With this trend comes also a growing awareness of the chances offered by digitisation. Digital data are not only more easily accessible, i.e., they can be read concurrently and non-locally and are searchble, but they can also be automatically processed, analysed, visualised, and mined, provided that suitable tools are available. Digitisation along with the development of suitable text analysis techniques thus has the potential to revolutionise humanities research.

This tutorial aims to provide an introduction to the fairly novel but rapidly growing field of Digital Humanities (DH), with a specific focus on techniques for text analysis and mining. The tutorial thus complements Michael Piotrowski's tutorial on Natural Language Processing for Historical Text. Digital data in the humanities and cultural heritage domains are highly heterogeneous, ranging from meta-data describing physical exhibits in museums (such as catalogue or field book entries), over digitised inscriptions and letters to fiction books, newspapers or historical documents. Just as varied as the text types are the applications for language processing.  Structure detection methods can be used to turn field books into easily searchable databases. Error detection can enforce consistency in meta-data. Semantic analysis can increase the interoperability between various data sources and provide support for sophisticated data querying. Extracting and visualising social networks of characters in narratives can be a useful tool in literature analysis. Event detection and linking can help historians analyse large amounts of historical documents. The tutorial will provide an overview of such applications and of the methods used.


Caroline Sporleder

A A A | Drucken Print | Impressum Impressum | Sitemap Sitemap | Kontakt Contact
zum Seitenanfangzum Seitenanfang