The Knowledge Graph Conference Icon
The Knowledge Graph Conference
  • 馃彔Home
  • 馃搮Events
  • 馃懁Members
  • 馃數Announcements
  • 馃數Ask
  • 馃數Ask The Ontologists
  • 馃數Events
  • 馃數Jobs
  • 馃數Promotions
  • 馃數Share
Powered by Tightknit
Share
Share

Exploring Datalog Engines for Efficient Data Management and Querying

Avatar of Henrik P.Henrik P.
路Feb 19, 2022 02:41 PM

Do you agree? https://www.linkedin.com/pulse/syntax-semantics-great-owl-hoax-jan-voskuil/ Instead of applying global, abstract and generic rules using OWL (danger, Will Robinson!), why not use a datalog engine to craft very specific rules, in order to a) simplify your queries, b) perform some classifications, c) create hash values for indexing, and d) move load from query time (many times) to data ingestion time (once), trading memory for performance? I remember being super excited when I first learned about the first ever availability of a linked data datastore with an integrated datalog/rules engine, in RDFox (not affiliated), and even more excited when I saw how it worked in real life when developing an airline style pricing, availability, and efficient selling search engine. And, boy, was it fast! (Feel free to reach out if you want to hear more)

2 comments

路 Sorted by Oldest
    • Avatar of Henrik P.
      Henrik P.
      路

      Hope this was the right channel - could not seem to find a channel dedicated to sharing and discussing articles like this in general, and specifically about reasoning and rules

      馃挴1
    • Avatar of Bob DuCharme
      Bob DuCharme
      路

      Here's one answer to your "why not" question: because while the OWL dialects are a mess of overlapping standards, at least they are specs from a standards body where anyone can look up what syntax is supposed to do what. Datalog is whatever a given implementer wants it to be. If you implement something with one datalog engine and then move it to another, how much rewriting do you have to do?

      馃憤2