The data contained in the archive is structured into sessions which typically include media (audio or video recordings, sometimes images) and a metadata description of the circumstances of the recording. Usually, also transcription and annotation files are supplied. It is possible to search the whole archive, a subarchive created by a DOBES project, or in fact, one can search across data contained under any subnode in the archive.
Metadata search
For an exploration of the archive in terms of language of the session, location, participant or other information about the recording session, the metadata descriptions can be searched. By right-clicking on any corpus node (e.g. the top node of the DOBES archive, the top node of a documentation project or any sub branch of the archive) and selecting “metadata search”, a search can be performed over the metadata descriptions. This is by default an unstructured, google-type search but an advanced search interface to search in specific metadata fields is also provided. All metadata descriptions are publicly accessible, no registration is required.
Annotation search
For users who are interested in finding instances of certain linguistic phenomena (e.g. ergative case), narratives or discussions pertaining to a certain topic (e.g. astronomy, the use of tobacco, etc. which might not be the main topic of the conversation and therefore not explicitly mentioned in the metadata description), it is possible to search the annotation content.
Similarly to the metadata search, the annotation content search is accessed via a right-click on any corpus node. The annotation search can be unstructured or structured and is identical to the search functionality that is available in the ELAN annotation tool. Note that the depth and structure of annotation differs widely across projects and within projects across different sessions. Users are only allowed to search in the annotation files that are accessible to them.
See also: access methods, browsing