]> The LOD2 Consortium partners bring the essential know-how and software, which is necessary to build the LOD2 Stack. In particular, we have considered existing state-of-the-art software components developed by the LOD2 members which are briefly introduced in the following paragraphs. This software is freely available under an Open Source GPL license. 2 Comprehensive Knowledge Archive Network (CKAN) CKAN is a registry or catalogue system for datasets or other "knowledge" resources. CKAN aims to make it easy to find, share and reuse open content and data, especially in ways that are machine automatable. Those familiar with software services like freshmeat, CPAN, or apt can think of CKAN as providing an analogous service for content and data. The CKAN software is used to run the CKAN service: <a href="http://www.ckan.net/">http://www.ckan.net/</a> and is also running the catalogue behind <a href="http://data.gov.uk/">http://data.gov.uk/</a> as well as more than 15 other sites around the world including in Germany <a href="http://offenedaten.de/">http://offenedaten.de/</a>, Canada <a href="http://datadotgc.ca">http://datadotgc.ca</a>, and France <a href="http://fr.ckan.net">http://fr.ckan.net</a>. CKAN is a registry or catalogue system for datasets or other "knowledge" resources. CKAN aims to make it easy to find, share and reuse open content and data, especially in ways that are machine automatable. D2R Server is a tool for publishing the content of relational databases on the <a href="http://www.w3.org/2001/sw/">Semantic Web</a>, a global information space consisting of <a href="http://www.w3.org/DesignIssues/LinkedData.html">linked data</a>. Data on the Semantic Web is modelled and represented in <a href="http://en.wikipedia.org/wiki/Resource_Description_Framework">RDF</a>. D2R Server uses a customizable <a href="http://www4.wiwiss.fu-berlin.de/bizer/d2rq/spec/#specification">D2RQ mapping</a> to map database content into this format, and allows the RDF data to be browsed and searched – the two main access paradigms to the Semantic Web. D2R Server's Linked Data interface makes RDF descriptions of individual resources available over the HTTP protocol. An RDF description can be retrieved simply by accessing the resource's URI over the Web. Using a Semantic Web browser like <a href="http://www.w3.org/2005/ajar/tab">Tabulator</a> (<a href="http://www.w3.org/2006/Talks/1019-tab-tbl/">slides</a>) or <a href="http://www4.wiwiss.fu-berlin.de/bizer/ng4j/disco/">Disco</a>, you can follow links from one resource to the next, surfing the Web of Data. The SPARQL interface enables applications to search and query the database using the <a href="http://www.w3.org/TR/rdf-sparql-query/">SPARQL</a> query language over the SPARQL protocol. A traditional HTML interface offers access to the familiar Web browsers. <img src="http://www4.wiwiss.fu-berlin.de/bizer/d2r-server/images/architecture.png" alt="D2R Server architecture diagram" /> Requests from the Web are rewritten into SQL queries via the mapping. This on-the-fly translation allows publishing of RDF from large live databases and eliminates the need for replicating the data into a dedicated RDF triple store. Read more about the interfaces offered by D2R Server, including example HTTP requests and responses, in the Technical Note <a href="publishing/">Publishing Databases on the Semantic Web</a>. D2R Server is a tool for publishing the content of relational databases on the <a href="http://www.w3.org/2001/sw/">Semantic Web</a>, a global information space consisting of <a href="http://www.w3.org/DesignIssues/LinkedData.html">linked data</a>. Data on the Semantic Web is modelled and represented in <a href="http://en.wikipedia.org/wiki/Resource_Description_Framework">RDF</a>. D2R Server uses a customizable <a href="http://www4.wiwiss.fu-berlin.de/bizer/d2rq/spec/#specification">D2RQ mapping</a> to map database content into this format, and allows the RDF data to be browsed and searched – the two main access paradigms to the Semantic Web. D2R Server's Linked Data interface makes RDF descriptions of individual resources available over the HTTP protocol. An RDF description can be retrieved simply by accessing the resource's URI over the Web. Using a Semantic Web browser like <a href="http://www.w3.org/2005/ajar/tab">Tabulator</a> (<a href="http://www.w3.org/2006/Talks/1019-tab-tbl/">slides</a>) or <a href="http://www4.wiwiss.fu-berlin.de/bizer/ng4j/disco/">Disco</a>, you can follow links from one resource to the next, surfing the Web of Data. The SPARQL interface enables applications to search and query the database using the <a href="http://www.w3.org/TR/rdf-sparql-query/">SPARQL</a> query language over the SPARQL protocol. A traditional HTML interface offers access to the familiar Web browsers. <img src="http://www4.wiwiss.fu-berlin.de/bizer/d2r-server/images/architecture.png" alt="D2R Server architecture diagram" /> Requests from the Web are rewritten into SQL queries via the mapping. This on-the-fly translation allows publishing of RDF from large live databases and eliminates the need for replicating the data into a dedicated RDF triple store. Read more about the interfaces offered by D2R Server, including example HTTP requests and responses, in the Technical Note <a href="publishing/">Publishing Databases on the Semantic Web</a>. D2R Server is a tool for publishing relational databases on the Semantic Web. It enables RDF and HTML browsers to navigate the content of the database, and allows applications to query the database using the SPARQL query language. DBpedia is a community effort to extract structured information from Wikipedia and to make this information available on the Web. It currently already contains a tremendous amount of valuable knowledge extracted from Wikipedia. The DBpedia knowledge base will be used for evaluation LOD2’s interlinking, fusing, aggregation and visualization components. The DBpedia multi-domain ontology will be used as background-knowledge for the LOD2 applications (WP7, WP8 and WP9), and as an alignment and annotation ontology for LOD in general. DBpedia is a community effort to extract structured information from Wikipedia and to make this information available on the Web. It currently already contains a tremendous amount of valuable knowledge extracted from Wikipedia. The DBpedia knowledge base will be used for evaluation LOD2’s interlinking, fusing, aggregation and visualization components. The DBpedia multi-domain ontology will be used as background-knowledge for the LOD2 applications (WP7, WP8 and WP9), and as an alignment and annotation ontology for LOD in general. The DL-Learner software learns concepts in Description Logics (DLs) from examples. Equivalently, it can be used to learn classes in OWL ontologies from selected objects. It extends Inductive Logic Programming to Descriptions Logics and the Semantic Web. The goal of DL-Learner is to provide a DL/OWL based machine learning tool to solve supervised learning tasks and support knowledge engineers in constructing knowledge and learning about the data they created. Purposes of Class Expression Learning: <ol> <li>Learn Definitions for Classes: Based on existing instances of an OWL class, DL-Learner can make suggestions for class definitions to be included as an owl:equivalentClass or rdfs:subClassOf Axiom. As the algorithm is biased towards short and human readable definitions, a knowledge engineer can be supported when editing the TBox of an ontology (see <a href="http://dl-learner.org/wiki/ProtegePlugin">Protege Plugin</a>).</li> <li>Find similar instances: DL-Learner's suggested class expressions can be used to find similar instances via retrieval (Concept definitions as search). Scalable methods allow the generation of recommendations on the fly, e.g. in a web scenario (see <a href="http://navigator.dbpedia.org/">DBpedia Navigator</a> – in experimental stage).</li> <li>Classify instances: The learned class descriptions can be used in a typical classification scenario, i.e. to decide for unknown instances whether they belong to a certain class. Common ILP benchmarks have been tested with DL-Learner. On the <a href="http://dl-learner.org/wiki/Carcinogenesis">Carcinogenesis page</a>, DL-Learner competes with other state-of-the-art ILP algorithms.</li> The DL-Learner software learns concepts in Description Logics (DLs) from examples. Equivalently, it can be used to learn classes in OWL ontologies from selected objects. It extends Inductive Logic Programming to Descriptions Logics and the Semantic Web. The goal of DL-Learner is to provide a DL/OWL based machine learning tool to solve supervised learning tasks and support knowledge engineers in constructing knowledge and learning about the data they created. Purposes of Class Expression Learning: <ol> <li>Learn Definitions for Classes: Based on existing instances of an OWL class, DL-Learner can make suggestions for class definitions to be included as an owl:equivalentClass or rdfs:subClassOf Axiom. As the algorithm is biased towards short and human readable definitions, a knowledge engineer can be supported when editing the TBox of an ontology (see <a href="http://dl-learner.org/wiki/ProtegePlugin">Protege Plugin</a>).</li> <li>Find similar instances: DL-Learner's suggested class expressions can be used to find similar instances via retrieval (Concept definitions as search). Scalable methods allow the generation of recommendations on the fly, e.g. in a web scenario (see <a href="http://navigator.dbpedia.org/">DBpedia Navigator</a> – in experimental stage).</li> <li>Classify instances: The learned class descriptions can be used in a typical classification scenario, i.e. to decide for unknown instances whether they belong to a certain class. Common ILP benchmarks have been tested with DL-Learner. On the <a href="http://dl-learner.org/wiki/Carcinogenesis">Carcinogenesis page</a>, DL-Learner competes with other state-of-the-art ILP algorithms.</li>   </ol> DL-Learner is a tool for supervised Machine Learning in OWL and Description Logics. It can learn concepts in Description Logics (DLs) from user-provided examples. Equivalently, it can be used to learn classes in OWL ontologies from selected objects. It extends Inductive Logic Programming to Descriptions Logics and the Semantic Web. The goal of DL-Learner is to provide a DL/OWL-based machine learning tool to solve supervised learning tasks and support knowledge engineers in constructing knowledge and learning about the data they created. MonetDB is an open-source high-performance database system that allows to store relational, XML and RDF data, downloadable from monetdb.cwi.nl. While being well-known for its columnar architecture and CPU-cache optimizing algorithms, the crucial aspect leveraged in the scope of this project is its unique run-time query optimization framework which provides a unique environment to crack the recursive-correlated-self-join queries caused by semantic web queries to triple stores. MonetDB is an open-source high-performance database system that allows to store relational, XML and RDF data, downloadable from monetdb.cwi.nl. While being well-known for its columnar architecture and CPU-cache optimizing algorithms, the crucial aspect leveraged in the scope of this project is its unique run-time query optimization framework which provides a unique environment to crack the recursive-correlated-self-join queries caused by semantic web queries to triple stores. OntoWiki is a tool providing support for agile, distributed knowledge engineering scenarios. It facilitates the visual presentation of a knowledge base as an information map, with different views on instance data. It enables intuitive authoring of semantic content, with an inline editing mode for editing RDF content, similar to WYSIWIG for text documents. OntoWiki provides sophisticated means for navigating, visualising and authoring of RDF-based Knowledge Bases. It serves and consumes Linked Data and comprises a comprehensive middleware API for building custom Semantic Web applications. We refer to it as a Wiki, since our focus is on simplicity, adaptability and collaboration. However, other than annotating text-based Wiki pages with a special syntax (as suggested by text-based Semantic Wiki approaches), Onto Wiki uses RDF in the first place to represent information. For human users, Onto Wiki allows to create different views on data, such as tabular representations or maps. For machine consumption it supports various RDF serialisations as well as RDFa, Linked Data and SPARQL interfaces. Since its introduction in 2006, the application has evolved into a framework for building Semantic Web applications and was recently updated to support the collaboration across multiple domains and application via Semantic Pingback and RDFauthor OntoWiki is a tool providing support for agile, distributed knowledge engineering scenarios. It facilitates the visual presentation of a knowledge base as an information map, with different views on instance data. It enables intuitive authoring of semantic content, with an inline editing mode for editing RDF content, similar to WYSIWIG for text documents. PoolParty is a thesaurus management system and a SKOS editor for the Semantic Web including text mining and linked data capabilities. The system helps to build and maintain multilingual thesauri providing an easy-to-use interface. PoolParty server provides semantic services to integrate semantic search or recommender systems into systems like CMS, DMS, CRM or Wikis. PoolParty is a thesaurus management system and a SKOS editor for the Semantic Web including text mining and linked data capabilities. The system helps to build and maintain multilingual thesauri providing an easy-to-use interface. PoolParty server provides semantic services to integrate semantic search or recommender systems into systems like CMS, DMS, CRM or Wikis. SemMF is a flexible framework for calculating semantic similarity between objects that are represented as arbitrary RDF graphs. The framework allows taxonomic and non-taxonomic concept matching techniques to be applied to selected object properties. Moreover, new concept matchers are easily integrated into SemMF by implementing a simple interface, thus making it applicable in a wide range of different use case scenarios. Framework Architecture: he Matching Engine takes as input a query object and a collection of resource objects to be matched against the query object. Both are represented in RDF and may utilize different schema vocabularies. If they use concepts from a common taxonomy, an RDFS or OWL representation of this taxonomy has to be provided. SemMF Engine is implemented in Java utilizing <a href="http://jena.sourceforge.net/">Jena2 Semantic Web Framework</a> for accessing and querying of resource and query graphs as well as the underlying taxonomies. <img src="http://semmf.ag-nbi.de/doc/img/semmf_overview.jpg"> Matching Description: In most cases not every object property is relevant for the similarity computation. For example, an object representing a certain product may contain manufacturer's phone number which may be irrelevant for comparing product's characteristics with customer's preferences (query object). Thus, each relevant property in the query RDF graph must be explicitly specified and mapped to the semantically corresponding property (i.e. holding the same kind of information, e.g. price information) in a resource RDF graph. Each mapping is assigned a <a href="http://semmf.ag-nbi.de/doc/matchers.html">concept matcher</a> implementing a certain matching technique. In a <a href="http://semmf.ag-nbi.de/doc/creatingMD.html">matching description</a> the importance of each object property can be specified by assigning it a certain weight . Moreover, properties to be matched can be grouped into thematic clusters, for example all properties describing technical specification of a product. The property clustering enables to sort the matching result by cluster similarities. The matching description is represented in RDF using <a href="http://semmf.ag-nbi.de/vocabulary/1.1/semmf.rdfs">SemMF vocabulary</a> provided with the framework. Matching Process: Inside each thematic cluster the Engine calculates the similarity between each query property and the corresponding resource property. These similarities are multiplied by the indicated weights and summed up yielding the cluster similarity. All cluster similarities, in turn, multiplied by the specified cluster weights yield the object similarity. However, if for a given query property value there is more than one semantically corresponding resource property value (e.g. a product may be available in different colors) the Engine chooses the one with the highest similarity. Matching Result:: The output of the Matching Engine is a ranking of objects by their similarity values. The Engine also provides a <a href="http://semmf.ag-nbi.de/doc/traversingMatchingResult.html">detailed description of the matching process</a> (i.e. object property values, similarity values for all clusters and for each single object property within a cluster, associated weights, etc.) which can be used to generate explanations for the calculated object similarity. SemMF is a flexible framework for calculating semantic similarity between objects that are represented as arbitrary RDF graphs. The framework allows taxonomic and non-taxonomic concept matching techniques to be applied to selected object properties. Moreover, new concept matchers are easily integrated into SemMF by implementing a simple interface, thus making it applicable in a wide range of different use case scenarios <a href="http://Sig.ma">http://Sig.ma</a>is a tool to explore and leverage the Web of Data. At any time, information in Sigma is likely to come from multiple, unrelated Web sites - potentially any web site that embeds information in RDF, RDFa or Microformats (standards for the Web of Data). Sig.ma can be used in 3 main ways: <ul> <li>As a Web of Data browser: start from any entity and then click to another from the resulting page. Remember you are browsing a “network of mashups”, quite a unique thing. It might be noisy but you can spot gems, e.g. interesting description differences in different sources.</li> <li>As an embeddable/linkable widget: create a Sigma, refine it and when you’re ready to paste it around in emails and twits or embed it on your blog. Sigmas are “data live”: if one of your selected sources updates its information, so will your Sigma be updated wherever it shows.</li> <li>As a semantic API: retrieve entity descriptions and specific properties. For example picture,phone@Giovanni Tummarello , ready to consume, in JSON, in RDF.</li> </ul> Sig.ma is a tool to explore and leverage the Web of Data. At any time, information in Sigma is likely to come from multiple, unrelated Websites – potentially any website that embeds information in RDF, RDFa or Microformats (standards for the Web of Data). Sig.ma is a semantic web browser as well as an embeddable widget and also provides a Semantic Web API. The Web of Data is built upon two simple ideas: First, to employ the RDF data model to publish structured data on the Web. Second, to set explicit <a href="http://www4.wiwiss.fu-berlin.de/bizer/pub/LinkedDataTutorial/#links">RDF links</a> between data items within different data sources. Background information about the Web of Data is found at the wiki pages of the <a href="http://esw.w3.org/topic/SweoIG/TaskForces/CommunityProjects/LinkingOpenData">W3C Linking Open Data community effort</a>, in the overview article <a href="http://tomheath.com/papers/bizer-heath-berners-lee-ijswis-linked-data.pdf">Linked Data - The Story So Far</a> and in the tutorial on <a href="http://www4.wiwiss.fu-berlin.de/bizer/pub/LinkedDataTutorial/">How to publish Linked Data on the Web</a>. The Silk Link Discovery Framework supports data publishers in accomplishing the second task. Using the declarative Silk - Link Specification Language (Silk-LSL), developers can specify which types of RDF links should be discovered between data sources as well as which conditions data items must fulfill in order to be interlinked. These link conditions may combine various similarity metrics and can take the graph around a data item into account, which is addressed using an RDF path language. Silk accesses the data sources that should be interlinked via the SPARQL protocol and can thus be used against local as well as remote SPARQL endpoints. The main features of the Silk link discovery engine are: <ul> <li>Open source link discovery framework, running on all major platforms</li> <li>Support of RDF link generation (owl:sameAs links as well as other types)</li> <li>Flexible, declarative language for specifying link conditions</li> <li>Employment in distributed environments (by accessing local and remote SPARQL endpoints)</li> <li>Usable in situations where terms from different vocabularies are mixed and where no consistent RDFS or OWL schemata exist</li> <li>Scalability and high performance through efficient data handling (speedup factor of 20 compared to Silk 0.2): <ul> <li>Reduction of network load by caching and reusing of SPARQL result sets</li> <li>Multi-threaded computation of the data item comparisons (3 million comparisons per minute on a Core2 Duo)</li> <li>Optional blocking of data items</li> </ul> </li> </ul> Silk is implemented in Scala running on the Java Virtual Machine. In order to run Silk, developers need to: <ol> <li>Have SPARQL access to the datasets that should be interlinked.</li> <li>Write a link specification as described in the <a href="http://www4.wiwiss.fu-berlin.de/bizer/silk/spec/">Silk - User Manual</a>. </li> <li>Install the Silk framework as described in the <a href="http://www4.wiwiss.fu-berlin.de/bizer/silk/spec/index.htm#usage">Installation and Usage</a> section of the manual.</li> </ol> The Silk Linking Framework supports data publishers in setting explicit RDF links between data items within different data sources. Using the declarative Silk - Link Specification Language (Silk-LSL), developers can specify which types of RDF links should be discovered between data sources as well as which conditions data items must fulfil in order to be interlinked. These link conditions may combine various similarity metrics and can take the graph around a data item into account, which is addressed using an RDF path language. Billions of pieces of metadata are on the Web today, with increasing uptake across the Internet from <a href="http://www.google.com/support/webmasters/bin/topic.py?hl=en&topic=21997" >search engines</a> to <a href="http://developers.facebook.com/docs/opengraph" >social sites </a> to <a href="http://data.gov.uk/" >governments </a> alike. The key technologies are <a href="http://www.w3.org/RDF/">RDF</a>, <a href="http://www.w3.org/TR/xhtml-rdfa-primer/">RDFa</a> and <a href="http://microformats.org/">Microformats</a>. Examples of such information types are contacts, events, social networks, web polls, reviews, and hundreds of other domain specific entities. Sindice is a state of the art infrastructure to process, consolidate and query the Web of Data. Sindice collates these billions of pieces of metadata into an coherent umbrella of <a href="/developers/welcome">functionalities and services</a>. For more information, visit our <a href="http://blog.sindice.com/">blog</a> or <a href="http://groups.google.com/group/sindice-dev">support group</a>. Sindice is a state of the art infrastructure to process, consolidate and query the Web of Data. Sindice collates these billions of pieces of metadata into an coherent umbrella of functionalities and services. Sparallax is a faceted browsing interface for SPARQL endpoints, based on Freebase Parallax. This demonstrator showcases the benefits of intelligent browsing of Semantic Web data and represents a good starting point for LOD2 interfaces developed in WP 5. Sparallax is a faceted browsing interface for SPARQL endpoints, based on Freebase Parallax. This demonstrator showcases the benefits of intelligent browsing of Semantic Web data and represents a good starting point for LOD2 interfaces developed in WP 5. Triplify provides a building block for the “semantification” of Web applications. Triplify is a small plugin for Web applications, which reveals the semantic structures encoded in relational databases by making database content available as RDF, JSON or Linked Data. Triplify is very lightweight: It consists of only a few files with less than 500 lines of code. For a typical Web application a configuration for Triplify can be created in less than one hour and if this Web application is deployed multiple times (as most open-source Web applications are), the configuration can be reused without modifications. Triplify makes Web applications easier mashable and lays the foundation for next-generation, semantics-based Web searches. Triplify provides a building block for the “semantification” of Web applications. As a plugin for Web applications, it reveals the semantic structures encoded in relational databases by making database content available as RDF, JSON or Linked Data. Triplify makes Web applications easier mashable and lays the foundation for next-generation, semantics-based Web searches. Virtuoso is an innovative industry standards compliant platform for native data, information, and knowledge management. It implements and supports a broad spectrum of query languages, data access interfaces, protocols, and data representation formats that includes: SQL, SPARQL, ODBC, JDBC, HTTP, WebDAV, XML, RDF, RDFa, and many more. In addition to its core data management capabilities, Virtuoso also delivers sophisticated Data Virtualization functionality enabling the construction of federated views -- which may or may not be materialized -- over heterogeneous RDBMS, Web Services, and Hypermedia data sources. The open-source edition of Virtuoso, which includes a scalable high-performance RDF Quad Store, will be the basis for the LOD2 Stack's knowledge store. Virtuoso is a knowledge store and virtualization platform that transparently integrates Data, Services, and Business Processes across the enterprise. Its product architecture enables it to deliver traditionally distinct server functionality within a single system offering along the following lines: Data Management & Integration (SQL, XML and EII), Application Integration (Web Services & SOA), Process Management & Integration (BPEL), Distributed Collaborative Applications. The open-source data integration server and the highly efficient and scalable RDF triple store implementation in Virtuoso will be the basis for the knowledge store component in the LOD2 Stack. The WIQA - Information Quality Assessment Framework is a set of software components for filtering information from the Web using a wide range of different filtering policies. The framework has been designed to fulfill the following requirements: <ul> <li> Flexible Representation of Information together with Quality-Related Meta-information. Information quality assessment may rely on a wide range of different quality indicators. Which quality indicators are relevant depends on the application domain and the quality dimensions to be assessed. Important quality indicators in the context of web-based information systems are provenance information, ratings, and background information about information providers. The WIQA framework uses Named Graphs [<a href="http://www4.wiwiss.fu-berlin.de/bizer/wiqa/#CaBiHaSt04">CaBiHaSt05</a>] as a flexible data model for representing information together with quality related meta-information.</li> <li>Support for different Information Filtering Policies. The relevancy of different quality dimensions and the metrics used to assess these dimensions depend on the application domain, the quality indicators available, the task at hand and the subjective preferences of the information consumer. Therefore, information consumers use a wide range of different information filtering policies in different situations. The WIQA framework allows different policies to be employed for filtering information. Policies are expressed using a declarative policy language and can combine context-, content- and rating-based assessment metrics.</li> <li>Explaining Filtering Decisions. The accuracy of assessment results is often uncertain due to the limited availability of quality indicators and the uncertain quality of the quality indicators themself. Therefore, the final subjective decision of an information consumer whether to trust or distrust assessment results depends on his understanding of the quality indicators and the assessment metrics that have been used in the assessment process. In order to support information consumers in their trust decision, the WIQA framework can generate detailed explanations about filtering decisions.</li> </ul> The Web Information Quality Assessment Framework is a set of software components that empowers information consumers to employ a wide range of different information quality assessment policies to filter information from the Web. Information providers on the Web have different levels of knowledge, different views of the world and different intensions. Thus, provided information may be wrong, biased, inconsistent or outdated. Before information from the Web is used to accomplish a specific task, its quality should be assessed according to task-specific criteria.