Semantic Web Approaches for Cultural Heritage Modelling

Franziska Anne-Katrin Pannach

University of Göttingen
Göttingen Centre for Digital Humanities
Germany

Venue:
Room: Blue 1
Information Technology Building
University of Pretoria
Hatfield Campus

Workshop Description

In the first half of the workshop, we will provide an overview about Semantic Web with a special focus on ontologies. The participants will then be introduced to the ontology description language OWL (Web Ontology Language) to create a foundation of understanding about how data is presented in Semantic Web contexts. The workshop aims to delineate how humanities research can benefit from a combination of methods from knowledge engineering and natural language processing. In the second part of the workshop, the users will be introduced to one specific example of an ontology that can illustrate how the forementioned connection can be realised. Together, we will discuss possible applications that focus on the participants’ areas of interest.

Our ontology is designed as a tool for annotating and investigating folktales in various languages. It follows the approach of Russian formalist Vladimir Propp [Propp, 1968]. According to Propp, Russian magic tales are made up of a limited set of 31 functions that can be used to describe a tale. These functions follow a strict sequential order. Propp also claims that all of Afanasyev’s [Afanasyev, 1916] tales he studied are of one type in regard to their structure, i. e. their sequence of narrative functions [Propp, 1968]. During the workshop, the participants will be introduced to Propp’s theory in more detail. We have applied this theory to a small number of English translation of African tales [Smith, 1989],[Greaves, 1990], [Scheub, 2005]. However, we wish to expand the ontology to tales in native African languages. Our ontology has been carefully modelled using a description logic foundation that defines the logical interaction between agents and Proppian functions.

Subsequently, it has been implemented in OWL and will be published as a webapplication in January 2019. It allows analyses on an abstract plot level and on the level of characters and their interaction. A small set of data properties has been defined to allow representing the functions, tales and publications within the annotation. We intentionally limited the number of data properties to ensure that the ontology remains a useful annotation tool. Our ontology also models concepts that illustrate the family relationship of character‘s in a story [Koleva et al., 2012]. While these additional concepts and relationships are only based on Propp in the sense that they appear in the functions, they are nevertheless crucial elements for the analysis of the tales. So far, ten Bantu tales in their English translation have been annotated by four different annotators for test purposes. Although only one of the annotators had previous annotating experience, little assistance was needed for populating the ontology, which is an indicator for the usability of the system.

The participants will use the tool in a hands-on session for annotation and query purposes. In addition, we will demonstrate the use of the free Protégé software [Musen, 2015] with which individualized ontologies can be easily designed and implemented. We believe this workshop can help us learn from the participants, e.g. in which way they find tools of this sort useful for their specific area of interest, which needs they have for applying Semantic Web approaches in their research, and how we can establish collaborations from scholars of Computer Science and Humanities.

Target Group

The workshop is targeted at scholars from the Humanities, especially from the field of Folkloristics, Literature, Media and Cultural Studies and Linguistics. Furthermore, ontologies have been used to create interesting applications for Historians, e.g. in the field of Heraldry [Hiltmann, 2015]. It can also be of interest for researchers and staff from Cultural Heritage institutions such as museums or libraries who want to broaden their knowledge about Digital Humanities, Semantic Web and alternative ways of publishing structural data.

Organizatorical Requirements

The workshop is aimed to cover a four-hours session, with two hours for the theoretical foundations, one hour of practical work on the ontology and an hour to exchange ideas about possible applications for the participants' research questions. The workshop can take place in a computer-lab with connection to the internet or in a classroom venue if participants bring their own devices. Since all the basic ideas of Semantic Web, ontologies and the use of the web application/Protégé will be covered, no prior knowledge of the topic is needed. However, we would like to warmly invite interested scholars who have worked with semantic web applications before to come and share their experiences.

References

[Afanasyev, 1916] Afanasyev, A. N. (1916). Russian Folk-Tales. E. P. Dutton

& Company, New York.

[Greaves, 1990] Greaves, N., editor (1990). When the Hippo Was Hairy. Lutterworth Press.

[Hiltmann, 2015] Hiltmann, T. (2015). Digital heraldry: Digitisation and dis-

semination of the European heraldic heritage.

2[Koleva et al., 2012] Koleva, N., Declerck, T., and Krieger, H.-U. (2012). An

ontology-based iterative text processing strategy for detecting and recognizing characters in folktales. In Meister, J. C., editor, Digital Humanities 2012

Conference Abstracts, pages 467–470. University of Hamburg, Hamburg University Press.

[Musen, 2015] Musen, M. A. (2015). The protégé project: a look back and a

look forward. AI Matters, 1(4):4–12.

[Propp, 1968] Propp, V. (1968). Morphology of the Folktale, volume 10. University of Texas Press.

[Scheub, 2005] Scheub, H., editor (2005). African Tales. Univ. of Wisconsin

Press.

[Smith, 1989] Smith, A. M., editor (1989). Children of Wax: African Folk Tales.

Canongate.

Workshop #3: Semantic Web Approaches for Cultural Heritage Modelling