2 Replies Latest reply on Apr 21, 2008 2:03 PM by 32601

    Semantic Technologies Product Performance - Recourses for UniProt

    32601
    32601 Apr 17, 2008 12:04 PM
      Hi, where I can find the scripts and the exact location of the data file(s) for above benchmark regarding the UniProt part? The URL:

      http://www.oracle.com/technology/tech/semantic_technologies/htdocs/performance.html

      Following the links in reference [2] gives me some options to download different files. I am a bit confused which one to use.

      I can’t find the presentation for reference [3] anywhere on the Internet.

      I’m not sure how the dataset is imported into a staging table for above example. Are you using SQLLoader to load the file (XML or N-Triple) into the staging table? When yes, would it be possible to get hold of the control file?

      As part of my MSc dissertation I’m using the new UniProt RDF project found under http://dev.isb-sib.ch/projects/uniprot-rdf/ and I would like to compare my results with above scenario.

      Many Thanks.
      • 1971 Views
      • Tags:
        • 1. Re: Semantic Technologies Product Performance - Recourses for UniProt
          545538
          545538 Apr 18, 2008 11:24 AM (in response to 32601)
          If you installed the demo files from the Oracle Database Examples media, a sample control file is available at $ORACLE_HOME/md/demo/network/rdf_demos/bulkload.ctl. You can modify and use this file.
          • 2. Re: Semantic Technologies Product Performance - Recourses for UniProt
            32601
            32601 Apr 21, 2008 2:03 PM (in response to 545538)
            Hi Aleksey,

            Thanks for the information, should have known it earlier, I did a more complicated load using Java and PL/SQL. However I’m not interested in the benchmark results as such, more in verifying my imported data with a reference set as part of my dissertation.

            It is not a technical issue, just to verify my results in regards to data quality. At the moment I can’t produce a proof that my imported N-Triples and the resulting RDF model(s) are correct.

            My evaluation would go along these lines:
            The data and scripts kindly provided by A produce Xa N-Triples where Ya are duplicates, etc… My approach produces Xb N-Triples where Yb are duplicates, etc …. After inferencing Za, Zb triples are created …. Following SQL statements are used to verify both approaches with each other ….

            If anybody is out there who is using UniProt, please help.

            Many Thanks
            OH
            MSc Database Professional