Closed hudajkhan closed 4 years ago
This exploration should help us discern which linked data sources could be used and what are the mechanics/processes for retrieving suggestions and relationships based on user input and/or catalog results.
Slides define some potential opportunities.
SMASH! experimentation included the following: Retrieving based on term matching (includes partial matches) against QA for LCNAF and LCSH; retrieving broader, narrower, and close matches for LCSH; retrieving contemporary authors from the author index set up for BAM; looking author and subject facet values for the query results and looking up against FAST in QA to retrieve URIs as well as broader, narrower, close info for FAST subjects; querying Wikidata based on LCNAF URIs to get description and images; querying ANNIF for LCSH suggestions based on keywords.
What are the possibilities for incorporating linked data/semantics/ontologies which takes into account user's query and/or catalog results and facets? If query parsing options are to be included or reviewed, what NLP/ML tools can help with this task?