Wordnet As a Backbone of Domain and Application Conceptualizations in Systems with Multimodal Data

Jacek Marciniak


Abstract
Information systems gathering big amounts of resources growing with time containing distinct modalities (text, audio, video, images, GIS) and aggregating content in various ways (modular e-learning modules, Web systems presenting cultural artefacts) require tools supporting content description. The subject of the description may be the topic and the characteristics of the content expressed by sets of attributes. To describe such resources one can just use some of existing indexing languages like thesauri, classification systems, domain and upper ontologies, terminologies or dictionaries. When appropriate language does not exist, it is necessary to build a new system, which will have to serve both experts who describe resources and non-experts who search through them. The solution presented in this paper used to resource description, allows experts to freely select words and expressions, which are organized in hierarchies of various nature, including that of domain and application character. This is based on the wordnet structure, which introduces a clear order for each of these groups due to its lexical nature. The paper presents two systems where such approach was applied: the E-archaeology.org e-learning content repository in which domain knowledge was integrated to describe content topics and the Hatch system gathering multimodal information about the archaeological site targeted at a wide audience, where application conceptualization was applied to describe the content by a set of attributes.
Anthology ID:
2020.mmw-1.5
Volume:
Proceedings of the LREC 2020 Workshop on Multimodal Wordnets (MMW2020)
Month:
May
Year:
2020
Address:
Marseille, France
Venues:
LREC | MMW | WS
SIG:
Publisher:
The European Language Resources Association (ELRA)
Note:
Pages:
25–32
Language:
English
URL:
https://www.aclweb.org/anthology/2020.mmw-1.5
DOI:
Bib Export formats:
BibTeX MODS XML EndNote
PDF:
http://aclanthology.lst.uni-saarland.de/2020.mmw-1.5.pdf