Page History

...

Theme: VIVO Data Integrity and Cleanup

Revisiting VIVO data integrity and cleanup, e.g., the UF nightly queries, work at Cornell to delete orphaned date-time intervals, filters on ingest to prevent unknown (often just

...

misspelled) properties from getting in the triple store, counts of the number of triples in each named graph as a way to detect unexpected problems, etc.

Ideas for discussion

• the many phases of data cleanup – before conversion to RDF, after conversion but before ingest, after ingest, as a consequence of updates, after deletions
• various approaches – filters for malformed data, filters for data not conforming to any known ontology, orphaned data checks, graph-based checks
• various tools – what exists, what is needed
• ideas and approaches that you may have developed

...