Posts Tagged ‘dbpedia’

h1

the italian political activism and the semantic web

September 20, 2009

Beppe Grillo

A couple of years ago, during his live show, the popular italian blogger and activist Beppe Grillo provided a quick demonstration about how the Web concretely realizes the “six degrees of separation”. The italian blogger, today a Web enthusiast, shown that it was possible to him to get in contact with someone very famous using a couple of different websites: imdb, Wikipedia and few others. Starting from a movie where he acted, he could reach the movie producer and the producer could be in contact with another actor due to previous work with this latter and so on. 
This demonstration consisted in a series of links that were opened leading to some Web pages containing information where extract the relationships that the showman wants to achieve.
This gig came back to my mind while I was thinking on how, what I call the “Linked Data Philosophy”, is impacting on the traditional Web and I imagined what Beppe Grillo could show nowadays.
Just the following, simple, trivial and short SPARQL query:
<insert here the SPARQL query>
Although Beppe is a great comedian it may be hard also for him making people laugh with this. But, the point here is not about laughs but about data: in this sense, the Web of Data is providing an outstanding and an extremely powerful way to access to incredible twine of machine readable interlinked data.
Recently, another nice and remarkable italian initiative appeared on the Web: OpenParlamento.it. It’s, mainly, a service where the Italian congressmen are displayed and they are positioned on a chart basing on the similarity of their votes on law proposals.
Ok. Cool. But how the Semantic Web could improve this stuff?
First of all, it would be very straightforward to provide a SPARQL endpoint providing some good RDF for this data. Like the following example:
<rdf:RDF>
<foaf:name>Mario Rossi</foaf:name>
<foaf:gender>male</foaf:gender>
<openp:politicalGroup rdf:resource=”http://openparlamento.it/groups/Democratic_Party”/&gt;
</rdf:Description>
</rdf:RDF>
where names, descriptions, political belonging and more are provided. Moreover a property called openp:similarity could be used to map closer congressmen, using the same information of the already cited chart. 
Secondly, all the information about congressmen are published on the official Italian chambers web site. Wrapping this data, OpenParlamento.it could provide an extremely exhaustive set of official information and, more important, links to DBpedia will be the key to get a full set of machine processable data also from other Linked Data clouds.
How to benefits from all of this? Apart the fact of employing a cutting-edge technology to syndicate data, everyone who wants link the data provided by OpenParlamento.it on his web pages can easily do it using RDFa. Like the follow example, where a fragment of an HTML page representing a news on the above congressman:
<div>
</div>
contains some RDFa linking that page to the OpenParlamento.it cloud.
With these technologies as a basis, a new breed of applications (like web crawlers, for those interested in SEO) will access and process these data in a new, fashionable and extremely powerful way.

A couple of years ago, during his live show, the popular italian blogger and activist Beppe Grillo provided a quick demonstration about how the Web concretely realizes the “six degrees of separation”. The italian blogger, today a Web enthusiast, shown that it was possible to him to get in contact with someone very famous using a couple of different websites: imdb, Wikipedia and few others. Starting from a movie where he acted, he could reach the movie producer and the producer could be in contact with another actor due to previous work with this latter and so on. 

This demonstration consisted in a series of links that were opened leading to some Web pages containing information where extract the relationships that the showman wants to achieve.

This gig came back to my mind while I was thinking on how, what I call the “Linked Data Philosophy”, is impacting on the traditional Web and I imagined what Beppe Grillo could show nowadays.

Just the following, simple, trivial and short SPARQL query:

construct {
    ?actor1 foaf:knows ?actor2
}
    where {
    ?movie dbpprop:starring ?actor1.
    ?movie dbpprop:starring ?actor2.
    ?movie a dbpedia-owl:Film.
    FILTER(?actor1 = <http://dbpedia.org/resource/Beppe_Grillo&gt;)
}

Although Beppe is a great comedian it may be hard also for him making people laugh with this. But, the point here is not about laughs but about data: in this sense, the Web of Data is providing an outstanding and an extremely powerful way to access to incredible twine of machine readable interlinked data.

Recently, another nice and remarkable italian initiative appeared on the Web: OpenParlamento.it. It’s, mainly, a service where the Italian congressmen are displayed and they are positioned on a chart basing on the similarity of their votes on law proposals.

Ok. Cool. But how the Semantic Web could improve this stuff?

First of all, it would be very straightforward to provide a SPARQL endpoint providing some good RDF for this data. Like the following example:

<rdf:RDF>
    <rdf:Description rdf:about=”http://openparlamento.it/senate/Mario_Rossi”&gt;
        <rdf:type rdf:resource=”http://openparlamento.it/ontology/Congressman”/&gt;
        <foaf:name>Mario Rossi</foaf:name>
        <foaf:gender>male</foaf:gender>
        <openp:politicalGroup
            rdf:resource=”http://openparlamento.it/groups/Democratic_Party”/&gt;
        <owl:sameas rdf:resource=”http://dbpedia.org/resource/Mario_Rossi”/&gt;
    </rdf:Description>
</rdf:RDF>

where names, descriptions, political belonging and more are provided. Moreover a property called openp:similarity could be used to map closer congressmen, using the same information of the already cited chart. 

Secondly, all the information about congressmen are published on the official Italian chambers web site. Wrapping this data, OpenParlamento.it could provide an extremely exhaustive set of official information and, more important, links to DBpedia will be the key to get a full set of machine processable data also from other Linked Data clouds.

How to benefits from all of this? Apart the fact of employing a cutting-edge technology to syndicate data, everyone who wants link the data provided by OpenParlamento.it on his web pages can easily do it using RDFa.

With these technologies as a basis, a new breed of applications (like web crawlers, for those interested in SEO) will access and process these data in a new, fashionable and extremely powerful way.

Is the time for those guys to embrace the Semantic Web , isn’t it?