User egon willighagen - Semantic Overflow

XSLT-like RDF conversion with SPARQL instead of XPath?

2010-07-18T11:13:24Z

Joerg pointed me at this FOAF visualizer. Inspecting the source code, I noted it uses XSLT and XPath to find patterns. Since my FOAF file is embedded as RDFa, it does not surprise me that a lot of information is not visualized.

That made me wonder: I know XSLT and XPath are XML oriented... but, it would be nice to have something like XSLT to create HTML using SPARQL queries...

So, that we can do something like:

<xsl:template match="SELECT ?p ?name WHERE { ?p rdf:type foaf:Person; foaf:name ?name }">
  <html:div class="person">
    Name: <xsl:value-of select="?name"/>
  </html:div>
</xsl:template>

What system allows me to write a XSLT-like template like this, and if it exists, can you give an example to perform the above task?

Who set up this website, and how can we reach him/her?

2009-11-10T21:49:26Z

I have been wanting to ask the 'owner' of this forum to think about a weekly 'featured' questions, but I can't actually find a link with some contact information. The oldest account is Community, but that one is hardly active. Andrew perhaps?

How to set up a clustered Virtuoso on Debian/Ubuntu?

2010-06-23T12:42:22Z

Cygri mentioned in reply to the question Distributed triple stores that Virtuoso can be set up in a clustered mode. How would one do this using the Virtuoso that comes with Debian or Ubuntu? Please provide detailed configuration scripts.

How do I use Jena in a Google App?

2009-11-02T20:08:59Z

Is there a tutorial explaining in detail how Jena can be installed and used as part of a Google App using the Google Plugin for Eclipse, and in particular what Jena jars need to be uploaded to ensure no class loading problems are encountered?

Is SNORQL still actively developed?

2010-05-13T12:03:08Z

SNORQL is a nice interface to SPARQL end points, and I am using it myself. I was informed today about this JavaScript SPARQL editor (story), and that made me wonder if SNORQL is still developed, and where the most recent and up-to-date source code can be found. Is SNORQL still developed, under which license, and who might be interested in mashing it up with this SPARQL editor?

Answer by Egon Willighagen for XSD to OWL importer

2010-03-28T17:42:08Z

You can use a XSLT stylesheet to make this transformation, translating xsd:element into owl:Class, and child relations into owl:ObjectProperty. A minimal example is:

<?xml version='1.0' encoding='utf-8'?>
<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
    xmlns:xsd="http://www.w3.org/2001/XMLSchema"
    xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
    xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#"
    xmlns:owl="http://www.w3.org/2002/07/owl#">

    <xsl:output method="xml" encoding="utf-8" indent="yes"/>

    <xsl:template match="//xsd:schema">
      <rdf:RDF>
        <xsl:for-each select=".//xsd:element">
          <xsl:apply-templates select="."/>
        </xsl:for-each>
      </rdf:RDF>
    </xsl:template>

    <xsl:template match="xsd:element">
      <xsl:if test="./@name">
        <xsl:element name="owl:Class">
          <xsl:attribute name="rdf:about">
            <xsl:value-of select="./@name"/>
          </xsl:attribute>
        </xsl:element>
        <xsl:for-each select=".//xsd:element">
          <xsl:element name="owl:ObjectProperty">
            <xsl:attribute name="rdf:about">
              has<xsl:value-of     select="./@ref"/>
            </xsl:attribute>
            <xsl:element name="rdf:Range">
              <xsl:attribute name="rdf:resource">
                 <xsl:value-of select="./@ref"/>
              </xsl:attribute>
            </xsl:element>
          </xsl:element>
        </xsl:for-each>
      </xsl:if>
    </xsl:template>

</xsl:stylesheet>

Answer by Egon Willighagen for Potential usage for Purifyr?

2010-03-17T09:18:56Z

The great advantage of RDF is that it separates data from presentation. If you would use XHTML/RDFa, the HTML would not need to be cleaned at all, as the data would be cleanly provided anyway. So, cleaning HTML is not important if you already provide the data itself cleanly. The latter is what the RDF community focuses on.

How to query a SPARQL end point for properties for multiple resources?

2010-03-15T08:46:09Z

SPARQL end points are quite suited for extracting large tables of data from one resource. I can eaily ask for all protein sequences for all object, or limit this to 5, with something like (see this blog for more details):

PREFIX chembl: <http://rdf.farmbio.uu.se/chembl/onto/#>
PREFIX blueobelisk: <http://www.blueobelisk.org/chemistryblogs/>

SELECT DISTINCT ?act ?ass ?mol ?inchi
WHERE {
  ?act chembl:type \"IC50\" ;
    chembl:onAssay ?ass;
    chembl:forMolecule ?mol;
  ?mol blueobelisk:inchi ?inchi.
  ?ass chembl:hasTarget <http://rdf.farmbio.uu.se/chembl/target/t10885> .
}

However, when I want to take this result set and go to another SPARQL end point to query for properties for all molecules from this result set, I could:

make queries against the second end point for each InChI separately
query the second end point for all molecules with InChIs, and do a merge client side.

Both seem inefficient.

Is there not a nicer way of taking the result set of one SPARQL query as input to another, to query a specific set of resources, than the two sketched solutions?

LaTeX listings-like support for RDF serialization in Notation3?

2010-03-14T16:58:10Z

LaTeX has the nice listings package for putting source code into documents, but it does not seem to have syntax highlighting for Notation3. What alternative can I use?

Answer by Egon Willighagen for dct: or dcterms:

2010-03-14T10:10:18Z

I am not aware of guidance for prefixes, and as explained below, there should not be such guidance.

It does not matter which prefix you use (there is no right prefix), as long as it is properly tied to the namespace. These specs use two prefixes (dc and dcterms), but are merely used as shortcuts to their two matching namespaces (http://purl.org/dc/elements/1.1/ and http://purl.org/dc/terms/).

It is important to note that namespaces do not easily conflict, but prefixes can. It is therefore important that the actual prefix used does not matter. That said, there is a trade off between length of the document (short prefix) and human-readability (longer prefix).

Answer by Egon Willighagen for Is there a namespace to describe mimetypes and encodings?

2010-03-14T10:31:02Z

The NEPOMUK Information Element Ontology has a :mimeType predicate. Is has a xsd:string as range; the domain might be a problem, as it is defined as a :InformatationElement which may not be compatible with your application.

Limiting results in SPARQL to subjects with only one predicate of a certain type?

2010-03-09T17:03:11Z

Given I have an RDF graph ?a ?pred1 ?b, ?b ?pred2 ?c. Now, every ?a only has one ?pred1, but there may exists multiple ?pred2 for ?b. For example, ?pred2 could be rdfs:label.

Now, for my purposes, I want to query with SPARQL for results ?a, ?b, ?c, but if and only if ?b has one ?pred2 to a ?c.

If I use the SPARQL:

SELECT ?a ?b ?c {
   ?a ?pred1 ?b.
   ?b ?pred2 ?c.
}

I get "?a ?b" duplicated for each "?b ?pred2 ?c". However, I like to restrict these results to just listing the "?a ?b ?c" of there is only one triple ?b ?pred2 ?c.

That is, given the RDF triples:

:fooA :pred1 :fooB. :fooB :pred2 :fooC1;
:barA :pred1 :barB. :barB :pred2 :barC1;
                    :barB :pred2 :barC2;

I only want as results for the SPARQL:

:fooA :fooB :fooC1

What do I need to add to my SPARQL to make this happen?

OWL Axiom for inference on resource URIs?

2010-02-20T11:12:47Z

Given the following RDF triples:

@prefix swan: <http://swan.mindinformatics.org/ontologies/1.2/citations/>.
@prefix foaf: <http://xmlns.com/foaf/0.1/>.

<urn:doi:10.1186/1758-2946-2-1> a swan:JournalArticle ;
  foaf:homepage <http://dx.doi.org/10.1186/1758-2946-2-1> ;
  swan:doi "10.1186/1758-2946-2-1" ;

There is clearly a strong interdependency. Is there a way to reduce this graph, and infer these three equivalent triples from each other, for example, with OWL Axioms? The complicated bit is here in that the URIs contain the DOI...

What Jena/ARQ/Pellet API methods to call for OWL validation?

2010-02-20T09:39:53Z

Sort of in reply to the questions Simple CLI useable OWL Reasoner and Validate against rdf vocabulary?, I have the following question about what Java methods to call to validate:

an OWL ontology internally, and against the ontologies it builds upon (TBox)
a set of triples that uses that OWL ontology (ABox)

I am looking for actual Java code; the Jena, ARQ, Pellet and related websites have quite a few code samples, but all around from just one angle, and not a list of calls one would like to make sequentially. Is the RDF valid? (Possible even after: is the XML well-formed?) Is the OWL ontology valid? (Possible split up in various aspects) Is the ontology internally consistent= Is the data using the ontology consistent? And whatever other validation is possible in between...

What is the Java code that you use to sequentially run these validation tests?

Is there a web service that allow me to run SPARQL against a XHTML+RDFa website?

2010-02-20T13:33:44Z

It would be cool if, like those RDFa extraction services, there would be a service that would allow running a SPARQL query against an arbitrary XHTML+RDFa website and return it in, for example, JSON. Is there such a public service?

How to create RDFa-powered N3 in the HTML output?

2010-02-18T08:55:18Z

Related to the earlier question on Marking up a table with RDFa, I am looking for a XHTML+RDFa code snippet that outputs a triple as N3, but uses RDFa to make it machine readable at the same time. I am not so much looking for ways to markup the N3 line itself, which can already be considered machine readable, but for a code snippet that uses N3 for the visual output and RDFa for machine readability.

I have gotten to the below snippet but that is both rather verbose and has unwanted duplication:

<pre xmlns:bibo="http://purl.org/ontology/bibo/"
     xmlns:bibocto="http://github.com/egonw/bibo-cto/"
     about="urn:doi:10.1021/ci025584y"
     rel="bibocto:cites"
     typeOf="bibo:Article">
@prefix bibo: &lt;http://purl.org/ontology/bibo/>.
@prefix bibocto: &lt;http://github.com/egonw/bibo-cto/>.

&lt;urn:doi:10.1186/1758-2946-2-1> a bibo:Article ;
  bibocto:cites <span about="urn:doi:10.1021/ci025584y">&lt;urn:doi:10.1021/ci025584y></span> .
</pre>

Does someone have a more efficient solution?

Answer by Egon Willighagen for Any applications or libraries that implement OpenDocument RDF Data?

2010-02-09T09:15:09Z

Strigi is a desktop search engine (e.g. used by the KDE desktop) and indexes RDF data for OpenOffice documents:

> rdfindexer semanticOverflow.odt

How to use existing ontologies with Semantic Media Wiki?

2010-02-03T08:35:09Z

Bioclipse recently started using Semantic MediaWiki (thanx to Samuel), which allows me to define new classes, predicates and the sorts. For example, we have a custom class for our git repositories.

I was wondering, though, if people are also using existing ontologies (e.g. Dublin Core) on semantic wikis? Do you just replicate the terms, or is there a more clever way of importing that existing ontology into the wiki?

Semantic LaTeX creating HTML+RDFa?

2009-12-04T18:49:39Z

There are quite a few questions and answers about how to create HTML-RDFa. But regarding writing my papers, I am very fond of LaTeX. Now, I mostly use this to create PDFs, which are known hamburgers (see the cow-hamburgers paradigm). But I could also convert the LaTeX to (X)HTML. So, that make me wonder if anyone knows about how to enrich my LaTeX with RDFa. Besides producing semantic papers, it would also work for semantic slide presentations, etc.

Has anyone already done something like that? Would anyone be interested in drafting a LaTeX-RDFa proposal?

Java library to convert a SPARQL query into its CONSTRUCT variant?

2010-01-13T13:52:32Z

Is there a Java library that allows me to parse a SPARQL query and that can convert that query into a SPARQL version which uses the CONSTRUCT command to return RDF? Or am I being silly and are there better ways to use SPARQL to return RDF?

How to speed up rdf:type foo:Bar SPARQL queries by Virtuoso tuning?

2010-01-15T19:32:50Z

I'm running SPARQL queries like:

prefix ch: <http://pele.farmbio.uu.se/chembl/?>

select count(?s) where {
  ?s a ch:Compound .
}

or any other 'a' query asking all Resources of a certain rdf:type, which is quite slow resulting in a time out... how can I get Virtuoso to be a bit speedier here? Can I index the store so that it more efficiently can return all resources rdf:type some:Foo?

Is there an open source, faceted browser based on SWT widgets?

2009-11-07T15:06:56Z

I am looking for a open source, faceted browser based on SWT widgets or Eclipse Rich Client Platform technologies. Is anyone aware of such functionality? Please indicate what license the code is available under.

Answer by Egon Willighagen for SKOS Relationships

2010-01-05T10:43:33Z

MetWare uses SKOS because the domain is rather fuzzily modeled, but we extend the skos:broader and skos:narrower predicates to add additional meaning.

<rdf:Property rdf:about="&metware;isFieldOf">
  <rdfs:subPropertyOf rdf:resource="&skos;broader"/>
  <rdfs:label>Is Field Of</rdfs:label>
  <rdfs:comment>
    Metware specific relation that indicates that the
    subject is a field of the object.
  </rdfs:comment>
</rdf:Property>

What Open Source framework can autogenerate HTML for SPARQL output?

2009-11-19T13:51:31Z

I am looking for an overview of solutions for Open Source web frameworks which use some HTML template mechanism and creates the content HTML directly from a SPARQL query I define? That is, I want to pass the web framework the SPARQL query and not have to care about creating the HTML GUI stuff for displaying the results. The framework would analyze the SPARQL itself, see how many fields it can expect and create HTML accordingly, allowing me to simply care about the SPARQL itself.

Preferably, the output would be Ajax-using XHTML+RDFa output, but at the moment I would settle for anything. The framework should allow something along the lines of this pseudo code:

sparql = "SELECT WHERE ...";

insert("header");
insert(query(remoteServerURL, sparql));
insert("footer");

The results should be more fancy than a simple table, and preferably even allow me to specify how certain data types should be visualized. For example, a object property of type xsd:date to be automatically converted into a calendar widget?

Answer by Egon Willighagen for Ontology for units of measure

2009-12-03T08:19:15Z

The Open Biomedical Ontologies also has an suitable ontology for units:

http://www.obofoundry.org/cgi-bin/detail.cgi?id=unit

The question is, are all mentioned unit ontologies compatible?

Answer by Egon Willighagen for Why an ontology is OWL full?

2009-12-01T14:01:16Z

Any OWL DL document is OWL Full too. OWL DL puts restrictions on the OWL to make it compatible with description logic.

Mapping Ontologies: are there alternatives to owl:sameAs ?

2009-11-19T08:56:04Z

With the numerous ontologies being set up (which I do not see as bad), we also see the need to map ontologies. I am aware of the use of upper-ontologies (like UMBEL) to help make such mappings, and I am aware that I can use owl:sameAs to make rdfs:Resource-s and rdfs:Class-es.

What I am looking for is a literature on good-practices in this area, and in particular about alternatives to the use of owl:sameAs. If the documentation has theoretical arguments for the various options, that would be rather interesting.

Is there such information available?

Are there (Semantic Web) applications that can make use of FOAF profile?

2009-11-08T13:20:21Z

Increasingly, browser will push some personal information to websites like Geolocation. Are there websites/browsers that use similar functionality using, for example, my FOAF profile?

For example, OpenID client websites could pull my OpenID account from my FOAF profile (see foaf:openid) and the browser could log in with having to ask me for my OpenID account...

But there is much more information in ones FOAF profile than just the OpenID, and other applications I am very much interested in too!

Answer by Egon Willighagen for Template system for RDF ?

2009-11-20T11:47:51Z

If your input is in an XML format, you could use XSLT for this.

But your favorite programming language could help too. This is a piece of Groovy code I wrote yesterday to convert a JSON format in N3:

http://chem-bla-ics.blogspot.com/2009/11/chempedia-rdf-1-sparql-end-point.html

What BIBO tools are you using?

2009-11-04T20:52:16Z

BIBO (http://bibliontology.com/) is a RDF ontology for the markup of references. It can complement BibTeX in allowing the reference data to be directly embedded in webpages with XHTML+RDFa. In the past, I patched JabRef for preliminary BIBO support, but am wondering what tools people are using, or know about, that support BIBO for editing, conversion to/from, etc.

Supporting Export into BIBO

JabRef (subset of BIBO 1.0)

Supporting Import from BIBO

Comment by Egon Willighagen

2010-07-26T08:42:21Z

Payam, please add the updated code as answer to your question, and then mark it as final answer.

Comment by Egon Willighagen

2010-07-21T05:29:19Z

Can you give an example on how I template would look like?

Comment by Egon Willighagen

2010-07-21T05:28:56Z

BTW, can you extend the above example, and see how I would use the current selector, to create a bit of HTML?

Comment by Egon Willighagen

2010-07-21T05:28:06Z

Perhaps you can start blogging about it, and build documentation bit by bit that way... with examples etc... if you do, please let me know the URL... it sounds rather interesting!

Comment by Egon Willighagen

2010-07-18T11:47:55Z

@Richard: <span rel="foaf:accountServiceHomepage" href="twitter.com/">Twitter</span>; should be fine too, right?

Comment by Egon Willighagen

2010-07-18T11:37:22Z

Do you have any particular OData in mind? From what I just read, it seems to may that OData can be much more unstructured...

Comment by Egon Willighagen

2010-07-18T11:07:06Z

Some explanation on why I should check out that talk would be useful...

Comment by Egon Willighagen

2010-06-09T09:20:03Z

I like CiTO very much, and it is currently aligned with SWAN.

Comment by Egon Willighagen

2010-04-17T06:36:14Z

Pierre, you should be able to delete your answer.

Comment by Egon Willighagen

2010-03-28T17:43:41Z

You should also have enough karma now to add links to the product homepages.

Comment by Egon Willighagen

2010-03-17T13:56:06Z

Can you point to an existing FOAF file that gives that behavior?

Comment by Egon Willighagen

2010-03-16T11:13:08Z

Thanx. Fixed it.

Comment by Egon Willighagen

2010-03-16T11:12:37Z

Interesting link, thanx!

Comment by Egon Willighagen

2010-03-14T17:24:27Z

Added comment on the lack of guidance and comment that that is a good thing.

Comment by Egon Willighagen

2010-03-14T10:19:05Z

Dear Steevc, did you write up a summary of the answers somewhere? Could you link to that perhaps?