Over the past week, our group’s focus has been on implementation work for OPeNDAP’s provenance browser. The approach we’ve taken relies upon three types of data:
- Sample provenance records for OPeNDAP’s server processing.
- Records describing OPeNDAP modules called by the server.
- Developers responsible for maintaining modules.
Each of these three types of data has been loaded into our project RDF triple store based on Prizms – an RPI-developed toolkit designed to aid in the generation of linked open data. For entries (2) and (3), Prizms generates RDF records based on the following web-accessible spreadsheets:
OPeNDAP Modules: https://docs.google.com/spreadsheet/ccc?key=0An84UEjofnaydFRrUF9YWk03Y3NHNjJqUEg0NUhUZXc&usp=drive_web#gid=0
OPeNDAP Developers: https://docs.google.com/spreadsheet/ccc?key=0ArTeDpS4-nUDdHdqU3ZFREQ4LTA1aGxJR2o2S1BFeWc&usp=drive_web#gid=0
OpenLink Virtuoso, our RDF triple store, is capable of returning results in varying formats (XML,CSV,JSON) which can in-turn be used to generate human-readable presentations. Using XML encodings of query results, we are now applying a set of XSLT transformations to generate description pages for various forms of content. Below are some screenshots with accompanying descriptions.