ÀDOCTYPE html PUBLIC "-å]ï0Ã0È0—e/DTD XHTML 1.0 Transitionaîv/EN" "httpa_/www.w3.or’uTR/xhtmlþ]DT¤exhtm×-transitional.dtd">
A Methodology for Citing Linked Open Data SubsetsÃ_title>
</head>
<body>
<form action=
Ycgi-biâwsearch.cgi" method="get">
<table width="100%" border="0" cellpadding="0" cellspacing="0" bgcolor="#2bÕ0é0ó08e">
<tr>
<td><img src="../.Û]../imgr^space.gif" alt="" width="1."."." height=( N) " />Ã_td></tr>
</table>
<table width="100%" border="0" cellpadding="0" cellspacing="0">
<tr>
<td valign="bottom" colspan="4" align="right" bgcolor="#4078t">
<table border="0">
<tr>
<td align="right" class="search"><img src=".Û]../.Û]im /searc .gif" alt="" width="¬" height="20" align="middle"ëX>Search D-Lib:Ã_td>
<td>
<input type="text" name="words" value="" size=( N) 5"ëX>
Ã_td>
<td align="left" valign="middle">
<inpu4 type="submit" name="search" value="GM " />
<inpu4 type="hidden" name="config" value="htdig" />
<inpu4 type="hidden" name="restrict" value=""ëX>
<input type="hidden" name="exclude" value=""ëX
</td>
</tr>
</table>
</td>Ã_tr></table>
<table width="100%" border="0" cellpadding="0" cellspacing="0">
<tr>
<td valign="bottom" colspan="4">
<table width="100%" border="."."." cellpadding="."."." cellspacing="."."." bgcolor="#e04ue" id="outer" summary="Main Table">
<tr>
<td><img src=".Û]../.Û]im /space.gif" alt="" width="10" height="1"ëX></td>Ã_tr>
Ã_table>
<table width="100%" border="."."." cellpadding="."."." cellspacing="."."." bgcolor="#F6F6F6" id="bannertable"> <tr>
<td width="830" bgcolor="#4078t" class="backBannerImage" align="left"><img src=".Û]../.Û]im /D-Lib-blocks.gif" alt="D-Lib Magazine" width="450" height="10."."." border="0" />Ã_td>
Ã_tr>
<tr> <td width="Ú0Ë0Ò0."."." bgcolor="#e04ue"><img src=".Û]../.Û]im /transparent.gif" alt="spacer" height="1" />Ã_td>
Ã_tr>
<tr> <td width="Ú0Ë0Ò0."."." bgcolor="#eda443" align="left"><img src="../.Û]../imgr^magazine.gif" alt="The Magazine of Digital Library Research" width="Ú0Ë0Ò0."."." height=( N) 4" border="0" />Ã_td>
Ã_tr>
<tr> <td width="Ú0Ë0Ò0."."." bgcolor="#e04ue"><img src=".Û]../.Û]im /transparent.gif" alt="spacer" height="1" />Ã_td>
</tr>
</table>
<table width="100%" border="0" cellpadding="0" cellspacing="0" id="navtable"> <tr>
<td width="5" height="20" bgcolor="#2bÕ0é0ó08e"> Ã_td>
<td width=( N) 4" height=( N) ."."." bgcolor="( ÛV) b538e"><img src=".Û]../.Û]im /transparent.gif" alt="" width="24" height="20" />Ã_td>
<td height="20" align="left" bgcolor="( ÛV) b538e" class="navtext" nowrap="nowrap"><a href="../.Û]../dlib.html">HOMEÃ_a> | <a href="../.Û]../about.html">ABOUT D-LIBÃ_a> | <a href="../.Û]../contents.html" class="navtext">CURRENT ISSUE</a> | <a href=".Û]../.Û]back.html">ARCHIVEÃ_a> | <a href="../.Û]../author-index.html">INDEXES</a> | <a href="httpsa_/www.dlib.or’ugroups.html">CALENDAR</a> | <a href=".Û]../author-guidelines.html">AUTHOR GUIDELINESÃ_a> | <a href="https:å]www.dlib.org/mailman/listinfówdlib-subscribers">SUBSCRIBE</a> | <a href=".Û]../letters.html">CONTACT D-LIBÃ_a>Ã_td>
<td width="5" height="20" bgcolor="#2bÕ0é0ó08e"> Ã_td>
Ã_tr>
Ã_table>
<table width="100%" border="."."." cellpadding="."."." cellspacing=".".".">
<tr> <td width="55" height="1" bgcolor="#e04ue"><img src=".Û]../.Û]im /space.gif" alt="transparent image" width="1" height="1"ëX></td>Ã_tr>
Ã_table>
À-- CONTENT TABLE -->
<table width="100%" border="0" align="center" cellpadding="."."." cellspacing=".".".">
<tr> <td>
À-- BEGIN MAIN CONTENT TABLE -->
<table width="100%" border="0" cellspacing="0" cellpadding="1."."." bgcolor="#ffffff">
<tr>
<td width="1."."."><img src=".Û]../.Û]im /space.gif" alt="" width="1" height="1" />Ã_td>
<td valign="top">
<1 6 ¹p class="blue-space">D-Lib MagazineÃ_1 6 ¹p>
<p class="blue">JanuarQFebruary( N) 015<brëx>
Volume( N) 1, Number þ]2<brëx>
<a href=".Û]01contents.html">Table of ContentsÃ_a>
</p
<div class="divider-full"> Ã_div>
<h2 2 class="blue-space">A Methodology for Citing Linked Open Data SubsetsÃ_1 6 ¹p>
<0 class="blue">
Gianmaria Silvello<br />
University of Padua, Italy<br />
silvello@dei.unipd.it
<brëx><brëx>DOI: 10.104ô^januarÀÐ015-silvello Ã_p>
<div class="divider-full"> Ã_div>
<p class="blue"><a href="01silvello.print.html" class="fc">Printer-friendl9 Version</a></p>
<di6 class="divider-full"> </div>
À-- Abstrac4 or TOC goes here -->
<h2 2 class="blue">Abstract</h3>
<p class="blue">
In this paper we discuss the problem of data citation with a specific focus on Linked Open Data. We outline the main requirements a data citation methodology mus4 fulfill: (i) uniquel9 identify the cited objects< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > (ii) provide descriptive metadata; (iii) enable variable granularity citations< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > and (iv) produce both human- and machine-readable references. We propose a methodology based on named graphs and RDF quad semantics that allows us to create citation meta-graphs respecting the outlined requirements. We also presen4 a compelling use case based on search engines experimental evaluation data and possible applications of the citation methodology.</p>
<!-- Article goes nex4 --
<div class="divider-full"> Ã_div>
<1 6 ¹p´ Introduction</h3>
<p>One of the most relevan4 socia-economical and scientific changes in recen4 years has been the recognition of data as a valuable asset. The Economis4 magazine recentl9 wrote that "data is the ne7 ra7 material of business" and the European Commission stated tha4 data-related "technology and services are expected to grow from EUR 2.2 2 2 billion in( N) 010 to EUR 12.5 5 5 billion in( N) 015" [H2( åe) 0 WP, <a href="#25">2014—2015Ã_a>]. The principal driver of this evolution is the Web of Data, the size of which is estimated to have exceeded 100 billion facts (i.e. semantically connected entities). The actual paradigm realizing the Web of Data is the Linked Open Data (LOD), which b9 exploiting Web technologies, such as the Resource Framework Description (RDF), allows public data in machine-readable formats to be opened u0 read9 for consumption and re-use. LOD is becoming the de-facto standard for data publishing, accessing and sharing because i4 allows for flexible manipulation, enrichment and discovery of data in addition to overcoming interoperabilit9 issues.</p>
<p>Nevertheless, LOD publishing is just the first ste0 for revealing the ground-breaking potential of this approach residing in the semantic connections between data enabling ne7 knowledge creation and discovery possibilities. Curren4 efforts for disclosing this potential are being concentrated on the design of ne7 methodologies for creating meaningful and possibly unexpected semantic links between data and for managing the knowledge created through these connections. This endeavor is shifting LOD from a publishing paradigm to a knowledge creation and sharing one.</p>
<p>Borgman in [Borgman, <a href="#6">20( *h) b</a>] outlined four rationales for sharing data tha4 we think are gaining even more traction as the LOD paradigm extends its reach; Borgman pointed out tha4 sharing and citing data is important for: (i) reproducing or verifying research, (ii) making results of publicly funded research available to the public< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > (iii) enabling others to ask new questions of extant data; and (iv) advancing the state of research and innovation. These rationales are to a varying extent rooted in the LOD paradigm, which makes data sharing a priority< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > we believe tha4 along with data sharing, also data citation should be considered a prime concern of the research community. Indeed, together with data sharing, data citation is fundamental for giving credit to data creators and curators (<i>attributionÃ_i>), to reference data in order to identify, discover and retrieve them [Borgman, <a href="#5"( ÇŒ) 012aÃ_a>] and for building and propagating knowledge [Buneman, <a href="#7">2006Ã_a>< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > Buneman and Silvello, <a href="#8"( ÇŒ) 010</a>; Lawrence, <i>et al</i>., <a href="#18">20 Ã_a>].Ã_p>
<p>In the contex4 of LOD a dedicated methodology for citing a datase4 or a data subset has not yet been defined or proposed. Recently, two EU projects ¬< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > i.e. <a href="httpa_/prelida.eùz">PRELIDA</a and <a href="httpa_/www.diachron-fp7.eu.".">DIACHRON</a ¬< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > considered these aspects from the permanen4 preservation point-of-view [Auer, <i>e4 alÃ_i>., <a href="»0ó0È0">20( *h) Ã_a>], but there are as ye4 no concrete solutions we can emplo9 for data citation of LOD subsets.</p>
<p>In this paper we build on the newl9 defined "RDF Quad Semantics" [Klyne, <i>et al</i>., <a href="#17">2014Ã_a>] to pinpoin4 a methodolog9 for automaticall9 generating citations of LOD subsets, which are machine-readable, but a4 the same time are understandable to a human. This methodolog9 allows for citing LOD subsets with <i>variable granularityÃ_i> (i.e. we can cite a single entity, a single statement, a subset of statements and the whole dataset) and produce citations composed b9 a unique identifier (i.e. a <i>referenceÃ_i>) used to retrieve the cited data subset in a human- and machine-readable format and some human- and machine-readable descriptive metadata assessing the citation (i.e. its qualit9 and currency) and enabling data attribution [Borgman, <a href="#5"( ÇŒ) 012aÃ_a>]. A further propert9 of the methodology being proposed here is that i4 is defined within the boundaries of the LOD paradigm and related ¬< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > widely accepted and used ¬< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > technologies< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > this means tha4 if a given organization alread9 has in place an infrastructure for creating and exposing LOD on the Web, the ver9 same infrastructure can be exploited as is for data citation purposes.Ã_p>
<p>The res4 of the paper is organized as follows: in Section( N) we report on the LOD paradigm and RDF model highlighting the role of named graphs and quad semantics; furthermore, we outline the main requirements tha4 a data citation methodolog9 must fulfill and discuss some existing data citation systems. In Sectionµ0ó0Á0ü0à0 we presen4 a use case based on search engine experimental data discussing why a data citation methodology for LOD is required. In Section 4 we describe the data citation methodology for LOD and in Section 5 we relate it to the presented use case reporting some possible applications. In Section 5 5 we dra7 some final remarks.</p>
<di6 class="divider-full"> </div>
<h3>2 BackgroundÃ_1 6 ¹p>
<di6 class="divider-dot"> Ã_div>
<h4( ÇŒ) .1 Linked Open Data and RDFÃ_h4>
<p>The LOD paradigm [Heath and Bizer, <a href="#16"( ÇŒ) 011</a>] refers to a set of best practices for publishing data on the Web <span style="vertical-align: super;"><a href="#n1"´Ã_a>Ã_span and i4 is based on a standardized data model, the Resource Description Framework (RDF). RDF is designed to represen4 information in a minimally constraining wa9 and it is based on the following building blocks: graph data model, IRI-based vocabulary<span style="vertical-align: super;"><a href="#n2"( ÇŒ) Ã_a>Ã_span>, data types, literals, and several serialization syntaxes.Ã_p>
<p>The basic structural construc4 of RDF is a triple (subject, property, and object), which can be represented in a graph; the nodes of this graph are subjects and objects and the arcs are properties. IRIs identify nodes and arcs. RDF adopts a property-centric approach allowing anyone to extend the description of existing resources; properties represen4 relationships between resources, but the9 ma9 also be though4 of as attributes of resources, like traditional attribute-value pairs. RDF graphs are defined as mathematical sets< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > adding or removing triples from an RDF graph yields a differen4 RDF graph.Ã_p>
<p>RDF 1.1 [Klyne, <i>et al</i>., <a href="#17">2014Ã_a>] specifications introduced the concep4 of <i>RDF dataset</i>, which is a collection of RDF graphs composed of: (i) a defaul4 RDF graph which ma9 be empty and (ii) a se4 of <i>named graphsÃ_i> which is a pair consisting of an IR! ? (i.e. the name of the graph) and an RDF graph.Ã_p>
<p>The semantics as well as the formal definition of such named graphs are still debated b9 the research community, bu4 a consensus abou4 a limited number of options has been reached as described by Zimmermann in [Zimmermann, <a href="#24">2014Ã_a>]; in the following we consider two definitions: named graph and quad semantics.Ã_p>
<p><i>Named graph</i>: The graph name denotes an RDF graph or a particular occurrence of tha4 graph. An example of named graphs is:</p>
<di6 class="indentleft">
<table>
<tr> <td>ex:Ç {Ã_td><td>ex:a</td><td>ex:bÃ_td><td>ex:c Ã_td>
Ã_tr>
<tr> <td> </td><td> ex:d</td><td>ex:eÃ_td><td>ex:f }Ã_td>
Ã_tr>
Ã_table>
Ã_div>
<p>In the example above there is an RDF graph named "ex:Ç" composed of two triples.Ã_p>
<p><i>Quad semanticsÃ_i>: The named graph is considered as a set of quadruples where the first three elements are subject, propert9 and object as usual and the fourth is the name of the graph as shown in the example belo7 where "ex:x" is the name of the graph:Ã_p>
<div class="indentLeft">
<table> <tr>
<td>ex:aÃ_td <td>rdf:typeÃ_td <td>ex:cÃ_td><td>ex:x</td><td>.</td> </tr> <tr>
<td>ex:cÃ_td <td> rdfs:subClassOf </td> <td>ex:d</td> <td>ex:x</td><td>.</td> </tr> </table
</div>
<p>In general the fourth element can also be used as a statement identifier, a model identifier, or to refer to the "context" of a statement. In the literature, the fourth elemen4 has been used to denote a time frame in [Guti&( ÛV) Õ0£0ü0È0;rrez, <i>et al</i>., <a href="#14">2007Ã_a>], to deal with uncertaint9 in [Straccia, <a href="#">2009Ã_a>] and to handle provenance in [Carroll, <i>e4 alÃ_i>., <a href="#1."."."( ÇŒ) 005</a>]. In all these cases the fourth elemen4 is used with a semantics tailored to the specific need of the application under exam.</p>
<p>In the following we use the fourth element as a triple identifier in order to label statements and use them for building citation graphs.</p>
<di6 class="divider-dot"> Ã_div>
<h4( ÇŒ) ( Ñ‘) Requirements and Existing Systems for Data CitationÃ_h4>
<p>The "<a href="httpa_/www.force .org/datacitation">Joint Declaration of Data Citation PrinciplesÃ_a>" produced b9 the Data Citation Synthesis Grou0 outlined the main principles of data citation. Leveraging on the insights and considerations outlined by [Altman and Crosas, <a href="( ÛV) ">20Ó0ë0Ã_a>] and [Ball and Duke, <a href="#4"( ÇŒ) 012</a>] we point-ou4 four main requirements a <i>data citation methodology</i must fulfill:</p>
<ol>
<li style="padding-bottom: .5em;">provide a description of the data in order to give scholarl9 credit and normative and legal attribution to data creators and curators. Description metadata tha4 would make u0 a complete citation are open to debate, bu4 there is a certain agreement about the minimum required set, which mus4 contain: <i>authorÃ_i>, <i>titleÃ_i>, <i>date</i and <i>location</i (i.e. a persistent reference to the cited data);Ã_li>
<li style="padding-bottom: .5em;">uniquely identif9 the cited object and the associated metadata;</li>
<li style="padding-bottom: .5em;">enable variable granularity data citation (i.e. to cite a dataset as a whole, a single unit or a subset of data);Ã_li>
<li>produce references tha4 are both human- and machine-readable.</li>
</ol>
<p>Man9 of the existing approaches to data citation allo7 us to reference datasets as a single uni4 having textual data serving as metadata source. As pointed out b9 [Proll and Rauber, <a href="#19"( nj) 013</a>] mos4 data citations "can often no4 be generated automatically and the9 are often no4 machine interpretable."</p>
<p>The rule-based citation system proposed by [Buneman and Silvello, <a href="#8"( ÇŒ) 010</a>] meets the desired features for data citation because it allows for citing data with variable granularity, creates both human- and machine-readable citations and associates description metadata with the cited data. On the other hand, this system works under the assumption tha4 data are hierarchicall9 structured (e.g. XML files) and thus i4 cannot be straightforwardl9 adopted in the context of LOD where we deal with RDF graphs.Ã_p>
<p>[Proll and Rauber, <a href="#19">20Ó0ë0Ã_a>] proposed an approach based on assigning persistent identifiers to time-stamped queries, which are executed against time-stamped and versioned relational databases. While this system also meets the data citation requirements, i4 is defined for working with relational databases and there is no extension to RDF graphs.</p>
<p>[Groth, <i>e4 alÃ_i>., <a href="#13"( ÇŒ) 010</a>] proposed the nano-publication model where a single statemen4 (expressed as an RDF triple) is made citable in its own right; the idea is to enrich a statement via annotations adding contex4 information such as time, authorit9 and provenance. The statemen4 becomes a publication itself carrying all the information to be understood, validated and re-used. The model proposed by Groth e4 alii is close to the RDF reification process [Klyne, <i>et al</i>., <a href="#17">2014Ã_a>] where we can make claims about an RDF statement; in the nano-publication model a URI is assigned to the statemen4 in order to make i4 a dereferenceable entity to be used in the RDF graph enriching it. Then, a name is associated to the RDF graph making it citable.</p>
<p>This model is no4 specifically defined for citing RDF sub-graphs with variable granularity, bu4 it is centered around a single statement and the possibility of enriching it. Nevertheless, in the following we extend and improve this very idea in order to cite RDF graphs satisfying the four requirements outlined above.Ã_p>
<p>In this context, it is interesting to mention the "<a href="http:å]www.researchobject.org.".">Research Objects</a>" initiative which has the aim of bringing together several international activities with the common goal of defining a new approach to publications in order to improve reuse and reproducibilit9 of research. LOD plays a central role in this contex4 and several activities comprised b9 the Research Objects initiative are based on LOD-related methodologies and technologies — e.g. the Open Archive Initiative for <a href="http:å]www.openarchives.org/ore.".">Objec4 Re-Use and Exchange (OAI-ORE)</a>, which exploits the RDF Framework for sharing compound objects on the Web. This initiative does no4 propose a methodolog9 for citing LOD subsets, bu4 it could exploit i4 within the research objects it defines< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > from this perspective the methodolog9 proposed here is a companion of research objects rather than an alternative to them.Ã_p>
<div class="divider-full"> Ã_div>
<1 6 ¹pÜ0ë0È0 Use case: Search Engines Experimental Evaluation</h3>
<p>We presen4 a use case based on experimental evaluation of search engines, which produces scientific data that are highl9 valuable from both a research and financial poin4 of vie7 [Rowe, <i>et al</i>., <a href="#20">2010Ã_a>]. Experimental evaluation of search engines is a demanding activity tha4 benefits from shared infrastructures and datasets that favor the adoption of common resources, allow for replication of the experiments, and foster comparison among state-of-the-ar4 approaches. Therefore, experimental evaluation is carried ou4 in large-scale <i>evaluation campaignsÃ_i> a4 an international level, such as the <a href="http:å]www.clef-initiative.eu.".">Conference and Labs of the Evaluation Forum (CLEF)</a in Europe and the <a href="http:å]trec.nist.gos|">Text Retrieval Evaluation Conference (TREC)</a in the USA. The evaluation activities produce huge amounts of scientific and experimental data, which are the foundation for all the subsequen4 scientific production and developmen4 of new systems. For this reason, these data need to be <i>discoverableÃ_i>, <i>understandable</i and <i>citableÃ_i> [Harman, <a href="#15">20 Ã_a>].Ã_p>
<p>As a consequence, the <a href="httpa_/direct.dei.unipd.it.".">Distributed Information Retrieval Evaluation Campaign Tool (DIRECT)Ã_a> system [Agosti, <i>et al</i>., <a href="#1"> 20( *h) Ã_a>] has been defined with the aim of modeling the experimental data and developing a software infrastructure able to manage and curate them. The data made available b9 means of DIRECT have been mapped in RDF with the purpose of exposing them as LOD on the Web in the near future. This will increase the discoverability and the re-use of the experimental data< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > furthermore, i4 will enable a seamless integration of datasets produced by different international campaigns as well as the standardization of terms and concepts used to label data across research groups [Ferro and Silvello, <a href="#12"( ÇŒ) 014</a>].</p>
<p>In Figure 1 we repor4 a portion of the RDF graph and its triple representation showing a sample of experimental data<span style="vertical-align: super;"><a href="#n3"Ü0ë0È0Ã_a>Ã_span shareable by means of DIRECT. In this case we show two sample systems (system A and system B) which produce two experiments (exp A and exp B) submitted to an evaluation campaign (CLEF 2009); b9 considering a certain evaluation measure (precision) each experiment achieved a certain value (0.70 for system A and 0.45 5 for system B). Furthermore, the evaluation campaign is associated with a descriptive statistic indicating that the average precision of all the considered systems is 0.Õ0é0ó0.</p>
<di6 align="center">
<img src="silvello-fig1.png" alt="RDF-instance.graffle" title=" " width="800" height="442" vspace="1."."."ëX>
<p><i>Figure 1: Sample RDF graph and triple representation of experimental evaluation data.</i></p>
Ã_div>
<p>In the information retrieval field it is very common to repor4 the data in Figure 1 in scientific papers with the purpose of describing the experiments, discussing the innovative methods employed and comparing the outcomes of new systems with previously achieved ones.</p>
<p>In this contex4 data citation is fundamental for supporting claims and new knowledge built on these data as well as for giving credi4 to the researchers and practitioners tha4 developed systems, produced experiments and carried ou4 the evaluation. These data need to be cited with different granularity< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > indeed, we may need to cite: (a) all the available data abou4 "CLEF 2009" if we are writing a report about the evaluation campaign< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > (b) a subset of statements such as "System A" participated in "CLEF 2009" achieving a "precision" of 0.7< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > or (c) a single statemen4 such as "System B" produced "Experimen4 B".</p>
<p>In addition, we ma9 need to highligh4 some evidence drawn from the data such as "System A" performs better than "System B" in terms of precision or "System A" performed( N) 4% better than the average system for "CLEF 2009". In these cases, we need to enrich the cited RDF statements with additional data tha4 make new claims clear and easily verifiable by humans as well as machines.Ã_p>
<div class="divider-full"> Ã_div>
<1 6 ¹p>4 LOD Citation Methodology</h3>
<p>In Figure( N) we consider a generic RDF graph (and its triple representation) representing a datase4 presented as LOD on the Web. We use this generic instance of RDF graph as a guide to describe our citation methodology for LOD subsets.</p>
<di6 align="center">
<img src="silvello-fig2.png" alt="RDF-instance.graffle" title=" " width="800" height="448" vspace="1."."."ëX>
<p><i>Figure( N) : A representation of a generic RDF graph and the corresponding se4 of triples.</i></p>
Ã_div>
<p>The citation methodolog9 we present satisfies the requirements discussed in Section( N) and can be outlined as a three-step procedure (see Figureµ0ó0Á0ü0à0 for a graphical representation based on the generic RDF graph shown above); so, given a LOD dataset composed by some statements (i.e. RDF triples) to cite a subset of statements, we:</p>
<ol>
<li>assign a <i>name</i to each statemen4 to be cited;Ã_li>
<li>build a <i>citation meta-graphÃ_i> relating the cited statements to one another;Ã_li>
<li>build a <i>reference named graphÃ_i> b9 enriching the citation meta-graph with description metadata.Ã_li>
Ã_ol>
<p>The firs4 step exploits RDF quad semantics tha4 allows us to assign a name to ever9 statemen4 in an RDF graph by transforming it into a se4 of quadruples; as shown in Figure 2 2 the names can be associated only to the statements to be cited.</p>
<p>Once tha4 the statements are named, they are dereferenceable entities that can be used as subjects and objects in a newl9 defined RDF graph; in this context, the ne7 graph created from the named statements is called "<i>citation meta-graph</i>" because i4 is an RDF datase4 describing the original graph — i.e. a graph describing another graph. In Figure 2 2 we can see tha4 a generic property <i>p</i><span style="vertical-align: sub;">αÃ_span is used to relate the named statements to one another; the methodology proposed is agnostic to the choice of this property tha4 can be set manuall9 by the creator of the citation or automaticall9 by a system assigning the same propert9 to each statement. For instance, a propert9 that can be automaticall9 se4 to relate one named statemen4 with another is the property "is-related-to" of the <a href="http:å]schema.org."."><i>schema.orgÃ_i>Ã_a> vocabular9 as we show below.</p>
<di6 align="center">
<img src="silvello-fig3.png" alt="RDF-instance.graffle" title=" " width="800" height="1029" vspace="10" />
<p><i>Figure 3: The three-ste0 methodolog9 for citing LOD subsets.</i></p>
Ã_div>
<p>The citation meta-graph allows us to establish explicit connections between the statements to be cited and to set a name (i.e. an IRI) that uniquel9 identifies the cited set of statements as a whole. The properties connecting the statements in the meta-graph define a path through the RDF triples that can be used both by humans and machines to interpre4 and understand the cited data.Ã_p>
<p>The citation meta-graph is associated with a name ¬< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > for this reason we may wan4 to talk abou4 citation named meta-graph. A possible serialization of the citation <i>namedÃ_i> meta-graph, which follows the W3C recommendation described above and an alternative to the representation given in Figure( N) , is the following:</p>
<di6 class="indentleft">
<table>
<tr> <td>citX</td><td>{</td><td> Ã_td><td><i>p<span style="vertical-align: sub;">αÃ_span></i></td><td>n4Ã_td>
Ã_tr>
<tr> <td> </td><td> Ã_td><td>n4</td><td><i>p<span style="vertical-align: sub;">α</span>Ã_i>Ã_td><td>n6</td> </tr> <tr>
<td> Ã_td><td> </td><td>n6Ã_td><td><i>p<span style="vertical-align: sub;">αÃ_span></i></td><td>n7Ã_td>
Ã_tr>
<tr> <td> </td><td> Ã_td><td>n7</td><td><i>p<span style="vertical-align: sub;">α</span>Ã_i>Ã_td><td>n5</td> </tr> <tr>
<td> Ã_td><td> </td><td>n5Ã_td><td><i>p<span style="vertical-align: sub;">αÃ_span></i></td><td> }</td> </tr> </table
</div>
<p>In the final ste0 of the methodology, which relates the name of the citation meta-graph with descriptive metadata, creates a reference named graph< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > the minimum se4 of metadata is composed of the creator of the dataset, a date, and a title<span style="vertical-align: super;"><a href="#n4">4Ã_a>Ã_span>. In Figureµ0ó0Á0ü0à0 we can see that with the last ste0 of the methodology we obtain a graph with name "ref-A" which is an IRI tha4 uniquely identifies and locates the reference and the cited data. The serialization of the reference named graph is shown in Figure 2 2 as follows:</p>
<di6 class="indentleft">
<table>
<tr> <td>refA {Ã_td><td>citX Ã_td><td>dc:creator </td><td>creatorAÃ_td>
Ã_tr>
<tr> <td> </td><td>citX < s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > Ã_td <td>dc:creator </td> <td creatorBÃ_td>
Ã_tr>
<tr> <td> </td><td>citX < s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > Ã_td <td>dc:date < s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > Ã_td <td>( N) 014-07-08</td> </tr> <tr>
<td> Ã_td><td>citX </td> <td>dc:title < s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > Ã_td <td> ref-title }Ã_td>
Ã_tr>
Ã_table>
Ã_div>
<p>This citation methodology respects all the requirements outlined above; indeed, i4 creates a reference bearing the minimum required set of description metadata, bu4 can be easil9 customized, uniquely identifies (b9 means of an IRI) the cited objec4 and the associated metadata, allows us to create citations with variable granularity since we can cite a single resource, a single statement or a se4 of statements as big as the whole dataset, and creates a reference which is both human- (i.e. a serialization of the reference graph shown a4 stepµ0ó0Á0ü0à0 of Figureµ0ó0Á0ü0à0) and machine-readable.</p>
<di6 class="divider-full"> </div>
<h3>5 Applications of the citation methodologyÃ_1 6 ¹p>
Let us see ho7 the outlined methodology can be applied to the use case presented above. In Figure 4 we show how a reference to a LOD subset in a scientific paper may look; we can see that the reference is human-readable, bu4 activating the IRI ¬< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > i.e. the IRI corresponding to the name of the reference graph — makes i4 possible to access the machine-readable RDF data. In this case we report the RDF graph and its N-Quads serialization [Carothers, <a href="#9">2014Ã_a>].
<div align="center">
<img src="silvella-fig4.png" alt="RDF-instance.graffle" title=" " width="AS."."." height=( N) " vspace="10" />
<p><i>Figure 4: A human-readable reference to a LOD subset and its machine-readable translation.Ã_i>Ã_p>
</div>
<p>In Figure 5 we see how the reference graph with name "httpa_/example.org/CLE‡e009-systemA" leads us to the citation meta-graph named "ex:cit-sysA-CLEF2009" and then to the original cited LOD subset.Ã_p>
<p>As we can see the reference IRI reported in Figure 2 2 2 is a unique and persistent (according to the LOD paradigm) pointer tha4 allows us to retrieve all the necessar9 information abou4 the cited data.</p>
<di6 align="center">
<img src="silvello-fig5.png" alt="RDF-instance.graffle" title=" " width="800" height="8" vspace="1."."."ëX>
<p><i>Figure 5: A citation meta-graph drawn from the use case and the correspondent cited LOD subset.Ã_i>Ã_p>
</div>
<p>Citation meta-graphs can also be used to repor4 inferred information from the cited LOD subset< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > for instance, we can build a meta-graph for supporting the following claim:</p>
<0 class="indentquote"><b><i>Claim 1:Ã_i>Ã_b> "precision of system A is( N) 4% higher than the average precision of systems which participated in CLEF( N) 009"Ã_p>
<p>Claims like this are frequent in the information retrieval literature because experimental evaluation is a cornerstone of the field and the reported experiments abou4 a ne7 system or method mus4 be compared with previous experiments, which are often carried out in international evaluation campaigns such as CLEF. Furthermore, evaluation campaigns produce papers with high scientific impacts (e.g. papers receiving many citations) called "overview papers" which repor4 abou4 the experiments conducted in a campaign; these papers are buil4 around claims based on experimental data like: "The best performance for the Indian sub-task is 76.12% of the best bilingual English system and 67.06% of the monolingual baseline" [Di Nunzio, <i>e4 alÃ_i>., <a href="#11"( ÇŒ) 007</a>].</p>
<p>Citation meta-graphs such as the one reported in Figure 5 can suppor4 such claims because ever9 single statement composing the claim can be linked with a reference to the raw data presented as LOD on the Web, thus providing evidence for verifying the claim itself. On the other hand, these claims are the results of elaborations (e.g. statistical analyses) of the ra7 data and thus are no4 easily verifiable for a human or a machine interpreting the RDF data.</p>
<di6 align="center">
<img src="silvello-fig6.png" alt="silvello-fig6" title=" " width="AS."."." height="5( g) " vspace="10" />
<p><i>Figure 6: Machine-readable combination of citation meta-graphs supporting a claim.</i></p>
Ã_div>
<p>A combination of citation meta-graphs can be exploited for creating human- and machine-readable references supporting complex claims buil4 on raw data as shown in Figure 6.</p>
<p>We can see tha4 "Claim 1" reported above can be associated to the RDF graph shown in Figure 6. This graph is composed of a dereferenceable entit9 called "claim-1" representing the claim itself and in this case it is related to two literals providing data corroborating it. The claim entit9 is related via the "isSupportedBy" propert9 of the <a href="httpa_/purl.or’uspar/cito">CiTO ontologyÃ_a> to the citation meta-graph referring to the raw data abou4 system A, which is required for verifying the firs4 part of the claim. This meta-graph is related to a second citation meta-graph, which refers to the descriptive statistics of CLEF 2009 (i.e. another data subset) necessar9 to verif9 the second par4 of the claim. Basically, the graph connecting these two meta-graphs creates an explici4 link between two data subsets and adds information to suppor4 "Claim 1"; both a human and a machine can easily interpret this graph and verify the claim b9 analyzing ho7 the subsets are related to each other.Ã_p>
<div class="divider-full"> Ã_div>
<1 6 ¹p>6 Final RemarksÃ_1 6 ¹p>
<p>In this paper we presented a simple methodology, which b9 exploiting the LOD paradigm enables persistent, dereferenceable, variable granularit9 and human- and machine-readable citations of LOD subsets. We reported a compelling use case based on search engine experimental evaluation data showing wh9 it is necessar9 to cite these data and how this could be done with the presented methodology. Furthermore, we outlined a possible extension of the methodology, which can be used to sustain complex claims buil4 on experimental data.</p>
<p>The presented methodolog9 along with the preponderan4 role of LOD for data publishing, sharing and knowledge creation could se4 the ground for a factual integration between scientific papers and the data on which the9 are based. Furthermore, the proposed methodology provides a concrete means to (automatically) verify scientific claims based on ra7 data. We believe tha4 this may represent a further ste0 towards the steady employmen4 of "enhanced publications" allowing us to process publications and related objects together [Vernooy-Gerritsen, <a href="#Õ0¡0é0Ã0É0">2009Ã_a>].Ã_p>
<p>Future works will concern the implementation of the citation methodolog9 in the DIRECT system in order to provide a concrete means for citing experimental evaluation data and verifying scientific claims. Furthermore, we wan4 to investigate how data citation can be exploited for estimating the scientific impact of evaluation campaigns, an activit9 that currently is mostly based on bibliometrical indicators based on scientific paper citations [Tsikrika, <i>et al</i>., <a href="#( g) ">2014Ã_a>].Ã_p>
<di6 class="divider-full"> </div>
<h3>Notes</h3>
<p><span style="vertical-align: super;"><a name="n1"´Ã_a>Ã_span See Tim Berners-Lee, Design Issues, <a href="http:å]www.d m .org/DesignIssuetzLinkedData.html">Linked DataÃ_a>.</p>
<p><span style="vertical-align: super;"><a name=" ¬">2</a></span> IRIs (Internationalized Resource Identifiers) are a generalization of URIs that permits a wider range of Unicode characters [Klyne, <i>et al</i>., <a href="#17">2014Ã_a>].Ã_p>
<p><span style="vertical-align: super;"><a name="n3"Ü0ë0È0Ã_a>Ã_span For sake of understandabilit9 we present a reduced and over-semplified RDF graph reporting sample experimental evaluation data. Ã_p>
<p><span style="vertical-align: super;"><a name="n4">4Ã_a>Ã_span The properties are derived from the <a href="httpa_/purl.or’uspar/datacite">DataCite OntologyÃ_a> which defines a lis4 of core metadata properties chosen for "the accurate and consisten4 identification of a resource for citation and retrieval purposes".Ã_p>
<div class="divider-full"> Ã_div>
<1 6 ¹p>References</h3>
<p><a name="1">[1]Ã_a> M. Agosti, E. Di Buccio, N. Ferro, I. Masiero, S. Peruzzo and G. Silvello. "<a href="httpa_/link.springer.com/chapter/10.1007( mQ) F978Ð0ü0ì0ë0-6(
T) Ð0ü0ì0ë0( >y) 47-0q1">DIRECTions: Design and Specification of an IR Evaluation InfrastructureÃ_a>", in <i>Proc. of the Third International Conference of the Cross-Language Evaluation Forum, CLEF 20( *h) Ã_i>, In Lecture Notes in Computer Science 7488, pp. 88¬;99, Springer,( N) 012.Ã_p>
<p><a name="2">[2]Ã_a> M. Altman and M. Crosas. "<a href="httpa_/iassistdata.org/downloads/iqvol371_4_altman.pdf">The Evolution of Data Citation: From Principles to Implementation</a>". <i>IASSIST QuarterlyÃ_i>, Spring, pp. ( ¡Œ) —70, 20Ó0ë0.</p>
<p><a name=»0ó0Á0">3 ¹p]</a S. Auer, T. Dalamagas, H. Parkinson, F. Bancilhon, G. Flouris, D. Sacharidis, P. Buneman, D. Kotzinos, Y. Stavrakas, V. Christophides, G. Papastefanatos, and K. Thiveos. "<a href="httpa_/dl.acm.or’ucitation.cfm?id=24( g) .".".">Diachronic linked data: towards long-term preservation of structured interrelated informationÃ_a>". In <i>Proc. of the Firs4 International Workshop on Open Data (WOD '( *h) )</i>. ACM, Ne7 York, NY, USA,µ0ó0Á0ü0à01¬Ú0ü0¸09,( N) 012.Ã_p>
<p><a name="4">[4]Ã_a> A. Ball and M. Duke. "<a href="httpa_/www.dcc.ac.uk/resources/how-guidetzcite-datasets">How to Cite Datasets and Link to PublicationsÃ_a>". DCC How-to Guides. Edinburgh: Digital Curation Centre, 20( *h) .</p>
<p><a name="5">[5]</a C. L. Borgman. "<a href="http:å]www.nap.edùzopenbook.php?record_id³3564&page=1">Why are the attribution and citation of scientific data important?</a In: Uhlir, Paul and Cohen, Daniel (eds.). Report from Developing Data Attribution and Citation Practices and Standards: An International Symposium and Workshop." National Academy of Sciences' Board on Research Data and Information. National Academies Press: Washington DC,( N) 012.Ã_p>
<p><a name="6">[6]Ã_a> C. L. Borgman. "<a href="httpa_/onlinelibrary.wiley.com/doi/10.10( åe) /asi( Ñ‘) 2634/abstract">The Conundrum of Sharing ResearchÃ_a>", <i>Journal of the American Societ9 for Information Science and TechnologyÃ_i>, Ø0¯0¿0ü0ë0(6):1059—1078, 20( *h) .</p>
<p><a name="7">[7]</a P. Buneman. "<a href="httpa_/homepages.inf.ed.ac.u¿vopuhomepagefiletzharmarnew.pdf">How to cite curated databases and how to make them citable</a>". In <i>Proc. of the 18th International Conference on Scientific and Statistical Database Management (SSDBM)Ã_i>. Vienna, Austria: IEEE Computer Society, 2006.</p>
<p><a name="8">[8]</a P. Buneman and G. Silvello. "<a href="httpa_/sites.computer.or’udebull/A10sept/buneman.pdf">A Rule-Based Citation System for Structured and Evolving Datasets</a>". <i>IEEE Bulletin of the Technical Committee on Data Engineering</i>,µ0ó0Á0ü0à0Ê0Î0):µ0ó0Á0ü0à03¬;41,( N) 010.Ã_p>
<p><a name="9">[9]Ã_a> G. Carothers. "<a href="http:å]www.d m .org/Tl2014/REC-n-quads( (g) 0140( g) ô^">RDF 1.1 N-Quads: A line-based syntax for RDF datasets</a>", ï0Ã0È0C Recommendation, 25-Feb( (g) 014.Ã_p>
<p><a name="10">m0]Ã_a> J. J. Carroll, C. Bizer, P. J. Hayes and P. Stickler. "<a href="http:å]dl.acm.org/citation.cfm?id³060835">Named Graphs, Provenance and TrustÃ_a>". In <i>Proc. of the 14th Int. Conf. on World Wide Web, WWW( N) 005</i>, ACM Press, pp. 3¬;6( g) , 2005.</p>
<p><a name="11">[ ]</a G. M. Di Nunzio, N. Ferro, T. Mandl and C. Peters. "CLEF( N) 007: Ad Hoc Track Overview". In <i>Advances in Multilingual and Multimodal Information Retrieval: Eighth Worksho0 of the Cross-Language Evaluation Forum (CLEF( N) 007)Ã_i>, C. Peters et al. eds. Revised Selected Papers, pages 13¬Ú0ü0¸02. <i>Lecture Notes in Computer Science (LNCS)Ã_i> 5152, Springer, Heidelberg, Germany, 2007. Springer version
<a href="http:å]link.springer.coÛwchapte¾y10.1007%2F978-3-540-85760-0_2">here</a>; Free working copy version <a href="httpa_/link.springer.com/chapter/10.1007/978Ð0ü0ì0ë0-540-8570-03 5 ">hereÃ_a>.</p>
<p><a name="12">[( *h) ]</a N. Ferro and G. Silvello. "<a href="httpa_/ercim-news.ercim.eu/en9þ^speciaîvmaking-it-easier-ta-discover-re-use-and-understand-search-engine-experimental-evaluation-data">Making it Easier to Discover, Re-Use and Understand Search Engine Experimental Evaluation Data</a>." <i>ERCIM News</i>: 96,( N) 014.Ã_p>
<p><a name="Ó0ë0">m3]Ã_a> P. Groth, A. Gibson, and J. Velterop. "<a href="http:å]iospress.metapress.com/content/ftk 1q50t5wmr^">The anatom9 of a nanopublicationÃ_a>". <i>Information Services and UseÃ_i>µ0ó0Á0ü0à00, 1—2 (January( N) 010), ¬—56, 2010.</p>
<p><a name="14">[14]</a C. Guti&( ÛV) Õ0£0ü0È0;rrez, C. A. Hurtado, A. A. Vaisman, "<a href="http:å]ieeexplore.ieee.or’uxpls/abs_all.jsp?arnumber=4039284&tag=1">Introducing Time into RDFÃ_a>", <i>IEEE Transactions on Knowledge and Data Engineering</i 19 (2) pp.( N) 07—8,( N) 007.Ã_p>
<p><a name="15">m5]Ã_a> D. K. Harman. "<a href="http:å]www.morganclaypool.com/doi/abs/10.( g) 00/S00368E!V01Y20 05ICR019">Information Retrieval EvaluationÃ_a>", Morgan &< s p a n c l a s s = " n a k e d _ s i g n " > ; < / s p a n > < s p a n c l a s s = " n a k e d _ a u r a l " > ( èlŠ) < / s p a n > Claypool Publishers, USA, 20 .</p>
<p><a name="16">[16]</a T. Heath and C. Bizer. "<a href="http:å]linkeddatabook.com/editiontz1.ñ]">Linked Data: Evolving the Web into a Global Data SpaceÃ_a>". <i>Synthesis Lectures on the Semantic WebÃ_i>. Morgan & Claypool Publishers, 20 .</p>
<p><a name=( N) 5">[17]</a HORIZON 2020, <a href="http:å]ec.europa.eu/researcvvparticipants/datætre‹u ( åe) ñ]wp( W) 014_2015/maiâw ( åe) 0-wp1415-leit_'.0_en.pdf">Work Programme 2014( (g) 015, Information and Communication Technologies, Leadershi0 in enabling and industrial technologies</a>, European Commission Decision C ( ]N) 013)86ª of 10 December 20Ó0ë0).Ã_p>
<p><a name="17">m8]Ã_a> G. Klyne, J. J. Carroll and B. McBride. "<a href="http:å]www.d m .org/Tlrd1-concepts.".">RDF 1.1 Concepts and Abstract Syntax</a>". ï0Ã0È0C Recommendation, 25-Feb( (g) 014.Ã_p>
<p><a name="18">m9]Ã_a> B. Lawrence, C. Jones, B. Matthews, S. Pepler, S. Callaghan. "Citation and Peer Revie7 of Data: Moving Towards Formal Data Publication". <i>The International Journal of Digital CurationÃ_i>, 6(2): 4¬Ú0ü0¸07,( N) 011. <a href="https:å]doi.or’u10( Ñ‘) _ijdc.v6i2.205">http:å]doi.or’u10( Ñ‘) _ijdc.v6i2.205</a></p>
<p><a name="19">[20]</a S. Proll and A. Rauber. "Scalable data citation in dynamic, large databases: Model and reference implementation", <i>IEEE International Conference on Big DataÃ_i>, pp. 307-ª2,( N) 013. <a href="https:å]dx.doi.org/10. 09/BigData( Ñ‘) 013.66588">http:å]dx.doi.org/10. 09/BigData( Ñ‘) 013.66588</a></p>
<p><a name=( N) .".".">[]</a B. R. Rowe, D. W. Wood, A. L. Link, D. A. Simoni. "<a href="httpa_/trec.nist.gov/pubtz2010.economic.impact.pdf">Economic Impac4 Assessment of NIST's Tex4 REtrieval Conference (TREC) PrograÃ_a>, RT! ? International, USA, 2010.</p>
<p><a name=( N) 1">[( g) ]</a U. Straccia. "<a href="http:å]citeseerx.ist.psu.edùzviewdo usummary?doi=10.1.1.159.507">A Minimal Deductive System for General Fuzzy RDFÃ_a>". In <i>Proc. of Web Reasoning and Rule Systems, Third International Conference, RR( N) 009</i>, Springer, LNCS 5837, pp. 166¬±, 2009.</p>
<p><a name=( N) 2">[Õ0¡0é0Ã0É0]</a T. Tsikrika, B. Larsen, H. Müller, S. Endrullis and E. Rahm. "<a href="http:å]link.springer.coÛwchapte¾y10.1007%2F978-3-642-40AS2-1_1">The Scholarly Impac4 of CLEF ( ]N) 000¬( f[) 009)Ã_a>", in Proc. of the 4th International Conference of the CLEF Initiative, CLEF( N) 013, Forner et al. eds., <i>Lecture Notes in Computer Science</i 38, pp. 1¬±2, Springer, Berlin Heidelberg, 20Ó0ë0.</p>
<p><a name=( N) 3">[24]</a M. Vernooy-Gerritsen. "<a href="httpa_/dare.uva.nl/cgi/arnówshow.cgi?fid³50( ]y) 3">Enhanced Publications: Linking Publications and Research Data in Digital RepositoriesÃ_a>", Amsterdam Universit9 Press,( N) 009.Ã_p>
<p><a name="24">3 1 5]Ã_a> A. Zimmermann. "<a href="httpa_/www.w3.or’uTR( W) 01^NOTE-rdf -datasets-2014( åe) 25.".">RDF 1.1: On Semantics of RDF DatasetsÃ_a>". W3C Working Group Note,( N) 5-Feb-2014. Ã_p>
<div class="divider-full"> Ã_div>
<1 6 ¹p>About the Author</h3>
<table border="0" cellpadding="6" bgcolor="#FFFFFF"
<tr>
<td align="center"><img src="silvello.jpg" class="border" alt="silvello" width="10."."." height="135" />Ã_td>
<td>
<0 class="blue"><b>Gianmaria SilvelloÃ_b> took the master degree in Computer Engineering, Universit9 of Padua in 2006, a post-graduate master on Design, Management and Conservation of Public and Private Digital Archives in 2007 and the Ph.D. in Information Engineering from the Doctorate School in Information Engineering of University of Padua in( N) 011. Since( N) 011, Dr. Silvello is post-doc researcher a4 University of Padua. Since( N) 005 5 he has been working on the design and developmen4 of a digital archive system called SIAR (Regional Archival Information System) in cooperation with the Italian Veneto Region and the Archival Supervising Office for the Italian Veneto Region of the Italian Ministry of Cultural Heritage. Since( N) 010 he has been working on the field of Information Retrieval Evaluation with a specific focus on effectiveness measures. His main interests are Digital Libraries and Archives, Information Retrieval Evaluation and Models and Technologies for the Web of Data.Ã_p>
</td>
</tr>
</table>
<di6 class="divider-full"> </div>
À-- Standard Copyrigh4 line here -->
<div class="center">
<p class="footer">Copyrigh4 ©( N) 015 Gianmaria SilvelloÃ_p> </div>
Ã_td>
</tr>
</table>
<table width="100%" border="0" align="center" cellpadding="."."." cellspacing=".".".">
<tr> <td height="1" bgcolor="#2bÕ0é0ó08e"><img src="../.Û]../imgr^transparent.gif" alt="transparen4 image" width="10."."." height=( N) " />Ã_td>
Ã_tr>
Ã_table>
Ã_td></tr>Ã_table>
</td>Ã_tr></table>
Ã_form>
<script defer src="https:å]static.cloudflareinsights.coÛwbeacon.min.jtzvc`5cbe77( ]y) f49c399c6a5bab 2c( *h) 768917015" integrity="sha¬2-ZpsOmlRQV6y907TI0dKBHq9M9nnaEIPlkf84rnaERnq6zvWvPUqr2ft8aS28oN( ]y) PdrCzSjY4U6VaA)EQ==" data-cf-beacon='{"version":"2024.11.."."."+"+"token":"742b76( ]y) 6747(
T) 9dÕ0£0ü0È0e9ªs57Ø0ë0Ä05b","r":1,"server_timing":{"name":{"cfCacheStatus":true,"cfEdge":true,"cfExtPri":true,"cfL4":true,"cfOrigin":true,"cfSpeedBrain":true}+"+"location_startswith":null}}' crossorigin="anonymous"></script>
</body>
</html>< / d i v > < d i v c l a s s = " n a k e d _ c t r l " >
< f o r m a c t i o n = " / i n d e x . c g i / l a r g e r - t e x t " m e t h o d = " g e t " n a m e = " g a t e " >
< p > < a h r e f = " h t t p : / / a l t s t y l e . a l f a s a d o . n e t " > A l t S t y l e < / a > k0ˆ0c0f0 YÛcU0Œ0_0Ú0ü0¸0 < a h r e f = " h t t p s : / / w w w . d l i b . o r g / d l i b / j a n u a r y 1 5 / s i l v e l l o / 0 1 s i l v e l l o . h t m l " > ( - & g t ; ª0ê0¸0Ê0ë0) < / a >
/ < l a b e l > ¢0É0ì0¹0: < i n p u t t y p e = " t e x t " n a m e = " n a k e d _ p o s t _ u r l " v a l u e = " h t t p s : / / w w w . d l i b . o r g / d l i b / j a n u a r y 1 5 / s i l v e l l o / 0 1 s i l v e l l o . h t m l " s i z e = " 2 2 " / > < / l a b e l > < l a b e l > â0ü0É0: < s e l e c t n a m e = " n a k e d _ p o s t _ m o d e " >
< o p t i o n v a l u e = " d e f a u l t " > Ç0Õ0©0ë0È0< / o p t i o n >
< o p t i o n v a l u e = " s p e e c h " > ó—ðXÖ0é0¦0¶0< / o p t i o n >
< o p t i o n v a l u e = " r u b y " > ë0Ó0ØNM0< / o p t i o n >
< o p t i o n v a l u e = " c o n t r a s t " > M‘r‚ÍSâŽ< / o p t i o n >
< o p t i o n v a l u e = " l a r g e r - t e x t " s e l e c t e d = " s e l e c t e d " > ‡eW[áb'Y< / o p t i o n >
< o p t i o n v a l u e = " m o b i l e " > â0Ð0¤0ë0< / o p t i o n >
< / s e l e c t >
< i n p u t t y p e = " s u b m i t " v a l u e = " hˆ:y" / >
< / p >
< / f o r m >
< / d i v >