QUIKK - Quadrilingual Infrastructure of Karstology Knowledge
1 Project goals
Given the prominence of Karst in the geography of Slovenia and Croatia it is important for experts to exchange knowledge and cooperate. This is only possible if the concepts, terms and definitions in the specialised domain are used unambiguously and experts can rely on a well-structured system of concepts to communicate with their peers. In a multilingual context, either when translating specialised texts into another language or when experts communicate in a non-mother language, efficient and faithful representation of the specialised contents becomes even more challenging.
The domain of karstology is indigenous to Slovenia and Croatia and is a well-researched field with rich terminology, consisting also of many local expressions referring to regionally specific phenomena. It is an interdisciplinary domain which can be studied from the geological, geographical, hydrological, environmental, historical and ethnological perspective. Dinaric Karst as a macroregion and landscape form has a large economic and cultural impact, it is the location of several natural parks and attracts numerous visitors to admire or explore its caves, engage in mountaineering, hiking or climbing, or simply observe nature and wildlife.
According to investigations performed so far (Grčić Simeunović 2014, Grčić Simenunović & Vintar 2015) Croatian and Slovene terms for typical geomorphological Karst structures are frequently borrowed into English and French, and foreign researchers often refer to works by native explorers of authentic Dinaric Karst. In view of this lively exchange of knowledge and terminology which, in contrast to many other domains where Croatian and Slovene are mainly passive recipients, is actively exported into English and French, there are strong reasons to build a structured knowledge base containing terms, definitions and hierarchical relations between concepts, for four languages: Croatian, Slovene, English and French.
The quadrilingual knowledge base will provide insight into conceptual structures and differences within the karst ontology, moreover it will demonstrate the multidimensionality and complexity of specialised concepts. It will be useful to geographers, geomorphologists, geologists, speleologists and other researchers of Karst phenomena. Using state-of-the-art computational tools to extract terminology from up-to-date corpora of specialised texts the project will create a multilingual hierarchical glossary reflecting authentic and contemporary use of specialised vocabulary in the karstology domain.
2 Project stages and expected results
a Extending Existing Corpora with Slovene
The Department of Romance Languages in cooperation with the Department of Geography at Zadar University have already started to compile a large comparable corpus of English, French and Croatian scientific and didactic texts pertaining to karstology. The corpus needs to be extended with Slovene texts and all four subcorpora need to be updated and balanced.
b Terminology Extraction and Validation
The existing LUIZ tool (Vintar 2010), which extracts terms from English and Slovene texts, will be adapted to French and Croatian. Term candidates will be extracted for all 4 languages and validated with the help of domain experts from both partner institutions.
c Building the knowledge base
Firstly, common karst concepts will be organised into a semantic network facilitating their analysis and comparison. Second, original (Croatian and Slovene) designations will be identified, including local or regional terms. Using the corpus and expert knowledge, definitions of concepts will be selected or created. The database will contain not only terms and definitions, but also relevant phraseology and knowledge-rich contexts. Finally, translation equivalents in English and French will be identified and added to the knowledge base.
The knowledge base will be created in the newly designed free tool Islovar provided by the Faculty of Arts, University of Ljubljana. The tool provides an online multi-user environment for collecting, managing and distributing data. Both partners will employ students of geography and languages/translation to help construct the knowledge base.
The main project result will be an online knowledge base of karstology concepts, terms, definitions and semantic relations containing translation equivalents in four languages: Croatian, Slovene, English and French.
We also envisage several joint publications following the analysis of comparable corpora and exploring concept dimensions across languages.