The document discusses the use of RDF (Resource Description Framework) and triplestores for reasoning about structured data, highlighting the significance of metadata in understanding complex relationships. It explains the organization of data as triples and the querying capabilities of SPARQL, which provides a powerful way to interact with various metadata schemas. Additionally, it mentions that while these technologies are innovative and part of the semantic web, they are still new and not widely utilized in production applications.

Similar to 21-RDF and triplestores in NOSql database.pptx

More from ajajkhan16

Recently uploaded

21-RDF and triplestores in NOSql database.pptx

• 1.
  RDF andtriplestoresCMSC 461MichaelWilson
• 2.
  Reasoning Relational databasesallow us to reasonabout data that is organized in a specificway Data that models specific relationships Data that is very cleanly structured What other reasoning methods areavailable to us?
• 3.
  Metadata “Data aboutdata” Data that describes other data Gives context Example metadata: Image EXIT data (geolocation, rotation,etc.) User statistics Last saved information in a file
• 4.
  What’s so important?The context that we gather frommetadata often allows us to understand amuch greater picture Can correlate and tie metadata together Calculate statistics on metadata Understand trends Infinite possibilities
• 5.
  The depth ofmetadata Many systems have their own way ofstoring metadata Database tables may be organized tohouse specific metadata This does not lend itself well to discoveringnew types of metadata Person may have age, DOB Later want to add new types (friends,Facebook ID, Twitter ID, etc.)
• 6.
  Metadata structures RDFResource Description Framework OWL Web Ontology Language Ontology – established vocabulary todescribe knowledge within a domain RDF is more widely used
• 7.
  Schemas RDF andother structured metadata formatsallow us to establish a common language todescribe different sorts of metadata We can make schemas that describe Social media Physical location Job details Moreover, we can tie them all to one subject Doesn’t require database reorganization
• 8.
  Why is thatcool? What this means is that we can tie anyarbitrary sets of data together with verylittle work on our part We make a schema that describes a newdomain, and staple that information ontoan existing subject
• 9.
  Triples Within theseschemas, data is conceptuallyorganized as <subject> <predicate> <object> Subject The subject of the expression Predicate The relationship between the subject and object Object The direct object of the expression These expressions are called “triples”
• 10.
  Triple examples Examples?
• 11.
  Storing triples Sincewe are often interesting in largeamounts of data, we need to think onhow to store these Triplestores Pretty obvious What do these give us over doingsomething like storing the information in adatabase?
• 12.
  Triplestore querying Triplestorescan also be queried SQL is more limited for the kinds of querieswe’d like to be able to make SPARQL The acronym stands for: SPARQL Protocol and RDF Query Language
• 13.
  SPARQL SPARQL isa SQL-like query language Allows us to query on the various schemaswe have assigned to our subjects SPARQL queries can look surprisinglyreadable
• 14.
  SPARQL examplePREFIX abc:<http://example.com/exampleOntology#>SELECT?capital ?countryWHERE {?x abc:cityname ?capital ;abc:isCapitalOf ?y .?y abc:countryname ?country ;abc:isInContinent abc:Africa .
• 15.
  Querying power UsingSPARQL, you can make extremelydeep, powerful queries and reason veryintuitively on the data present in atriplestore Organizing data this way allowscomputers to actually be able to reasonon data as well
• 16.
  Caveats All thistech is SUPER new All tied very heavily into the Semantic Web Basically introduce a system like this into theweb at large Metadata stored about web pages,computers can reason about them Much of this is a moving target Not a whole lot of production applicationsusing this stuff yet
• 17.
  Tools There area few triplestore servers andother tools you can use Jena Apache project Framework that allows for Semantic Webconcepts to be employed Can query using SPARQL Jena can use Postgres in the background
• 18.
  More tools RDFLibhttps://github.com/RDFLib Python library for RDF Can run entirely in memory Good for experimentation purposes andmore