Sharing linguistic multi-media resources at different complexity levels

  • Jacquelijn Ringersma, Max Planck Institute for Psycholinguistics, The Netherlands
  • Paul Trisbeek, Max Planck Institute for Psycholinguistics, The Netherlands
  • During the last decade much progress has been made in documenting endangered languages. Documentation results in sets of multi-media resources, consisting of audio recordings to represent spoken language and video recordings adding information on the socio-cultural environment and non-verbal communication. Primary data is complemented with transcriptions, translations etc.
    The MPI for Psycholinguistics houses an archive for linguistic resources, organized with rich metadata descriptions. Users can navigate the data by metadata search and/or annotation content search. The archive performs a major role in sharing resources between researchers.
    For non-professional users, searching archives for specific information can be complex because queries require advanced knowledge. Geographic navigation could be very powerful, since we may assume that these users will select resources according to language names or geographic areas. We provide a Google Earth overlay for the integrated presentation and sharing of archived resources.
    For speech communities, words are keys to describe relevant parts of their life and traditions. LEXUS is a web-based tool, allowing the creation of on-line lexica. Archived linguistic resources can be linked to the lexicon to provide extra information on the meaning of words. ViCoS, allows the linking of words to the concepts they denote, the creation and visualization of conceptual spaces, and the navigation from lexical to conceptual space. ViCoS can thus be used to create multi-dimensional concept networks that represent the richness of indigenous knowledge, anchored in the words used to describe it. This conceptual knowledge space facilitates sharing linguistic resources between researchers and the speech community.