In recent years Salzburg NewMediaLab, under the co-ordination of Salzburg Research, investigated research-related issues in the area of the “Linked Media Principles”, and developed an open reference implementation, the “Linked Media Framework“. At the same time a working group of the W3C prepared specification for a linked data platform, whose properties have not only been covered by the Linked Media Framework, but exceeded the specification. Therefore, the core of the Linked Media Framework was extracted and started as incubator project “Marmotta” at the Apache Foundation. This procedure guarantees legal certainty and support for the development and maintenance of the developer community and promotes sustainable development and evolution of open source software.

Apache Marmotta was launched in December 2012 as an incubating project and became a top level Apache project in November 2013. This is a recognition of years of intensive work in various research projects, which was pronounced by one of the largest and most important open source communities worldwide. In the so-called “incubation period” , the project must prove to meet the strict quality guidelines of the Foundation. This concerns both the copyright and license provisions of source code and libraries used, as well as dealing with the community and new developers. By enlarging and diversifying the developer team the software is provided on a stable, self – supporting basis.

The open software is already in use: The Red Bull Media House uses the principles for search and navigation in the central content pool. The Department of Documentation and Archives of the Austrian Broadcasting Company (ORF) is supported in the annotation of documentary contributions. The newspaper “Salzburger Nachrichten” use Marmotta for search and recommendations on salzburg.com involving a company’s internal knowledge base.

Further information: http://marmotta.apache.org

In its research program, Salzburg NewMediaLab extended the Linked Data Principles of the World Wide Web Consortium (W3C) to multimedia content (videos, photos, graphics, etc.), while at the same time realising W3C’s vision of the Read-Write-Web. The concept developed in Salzburg was called „Linked Media“, because the intention was to interlink structured and unstructured data and media content (or media fragments). The Linked Media concept enables organisations with content and media resources, to master the flood of data by semantic integration: enterprise data and content pools are linked to the Web of Data, allowing innovative solutions for search, navigation, annotation and recommendation.

In addition to the conceptual investigation of the „Linked Media Principles“, Salzburg NewMediaLab also came up with an open source reference implementation called the „Linked Media Framework“. Already at the time of the development of the framework, a working group at W3C published the specification of a generic “Linked Data Platform” (LDP) in response to the requirements of the growing Linked Data community. Since the Linked Media Framework not only covered the functional properties of this specification, but clearly exceeded the specification, the core of the Framework was extracted used to start an incubation project called “Apache Marmotta” (in English: “Marmot”) at the Apache Software Foundation (www.apache.org). The decision to move to the Apache Software Foundation was motivated by the increased awareness of industry that community based open source development is a desirable factor for product stability and business sustainability.

Apache Marmotta (http://marmotta.incubator.apache.org) was incubated in December 2012, published its 3.0 release in June 2013 and is ready to graduate and become an Apache top level project in the last quarter of 2013.

Appendices:

• SuccessStory: Ein Murmeltier gegen die Datenflut (DE)
• SuccessStory: A Marmot against the Flood of Data (EN)

From the  release note:

The team at Salzburg Research is happy to announce the final release of the Linked Media Framework, 3.0.0 version.
As usual, the project can be downloaded in different ways:

• LMF 3.0.0 WAR (deployable in any JavaEE Servlet container):
  http://code.google.com/p/lmf/downloads/detail?name=LMF-3.0.0.war
• LMF 3.0.0 Installer:
  http://code.google.com/p/lmf/downloads/detail?name=lmf-installer-3.0.0.jar
• Maven artifacts available from the Nexus

Major changes compared to 2.6.x:

• Bundled Apache Marmotta 3.0.0-incubating, providing Linked Data server capabilities and the KiWi triple store
• Enhancer updated to the latest version of Apache Stanbol EntityHub
• Semantic Search updated to Apache Solr 4.2.1
• Many other minor bug fixes and improvements

This is the first version bundling Apache Marmotta, and the development
will continue anyway there:  http://marmotta.incubator.apache.org

Download and availability

The LMF release is available at the LMF project website at:

• Project website: http://code.google.com/p/lmf/ resp. https://www.newmedialab.at/LMF
• standalone installer and deployable WAR: http://code.google.com/p/lmf/downloads/list
• Maven artifacts: (incl. Archetype): http://devel.kiwi-project.eu:8080/nexus
• Source code: http://code.google.com/p/lmf/source/browse/?name=lmf-parent-2.6.0

Major changes compared to 2.4.x:

• rewritten triplestore (including versioning, rule-based reasoner, transaction support) with considerably higher performance
• modularized LDPath, Linked Data Client and Linked Data Cache (so it can be used by other projects independently of the LMF)
• changed RDF API to Sesame (so it should be easy to exchange the underlying triple store)
• new administration interface
• started implementation of the upcoming Linked Data Platform specification
• many minor bugfixes and improvements and dependency cleanups

Outlook

This release is a milestone towards Apache Marmotta (incubating). Starting from now on, the Linked Data parts of the source code will be contributed to the Apache Software Foundation. The Linked Media Framework will continue to exist as a bundling of different technologies (including Linked Data Platform, Semantic Search, and Content Enhancement).

Apache Stanbol was founded in November 2010 by the community initiated by the EU research project IKS. It was the result of an ongoing discussion about how to ensure that the results, especially the developed software, of IKS would be available to vendors of content management systems (CMS) after the project’s official funding period ends.

Find the full story about Apache Stanbol and its graduation on the IKS-Website.

The highlights of the scientific-technical report are listed below (quote from part A of the report):

Research Achievements: Multimedia Annotation based on the Linked Media Principles

• Investigation of Linked Data Querying with LDPath and Simulation Unification
• Issue on “Smart Annotations” in the series of Linked Media Lab Reports
• Design of Linked Media interaction patterns
• Organisation of WWW Workshop “Semantic Web Collaborative Spaces”
• Along with a total of 36 scientific publications

Development Achievements: Multimedia Annotation based on the LMF

• Release of two versions (2.0 and 2.2) of the Linked Media Framework (LMF) as well as an One-Click-Installer
• Extension of the developer community by establishing an Open Source platform
• Integration of Apache Stanbol for content / media analysis
• Integration with Google Refine for triplification

Applied Research: Exploitation of Semantic Search and Multimedia Annotation

• Productive application of the Linked Media Principles in exploitation scenario „News and Information Platform“; implementation of the concepts developed in exploitation scenario “Sports Events” in the new release of the Red Bull Content Pool.
• More generally, already after the first two years of the K-Project, we identify a considerable impact of the research conducted at SNML-TNG on the partners’ product and process innovation. While the content partners started to include the Linked Media Principles in the plans for the design of their information and content platforms and partly already implemented solutions based on our approach, the technology partners integrated solutions based on the Linked Media Principles into their software platforms, partly also started marketing and awareness activities to prepare their customers and the respective market sectors for the Linked Media technology paradigm.

[sorry no English translation available for the rest of the article]

Gemeinsam mit den Kolleginnen und Kollegen vom IKS-Projekt wird am Stand der Salzburg Research unter anderen die Integration des Linked Media Frameworks (Release 2.2.0) mit Apache Stanbol vorgestellt (s. dazu auch den Blog-Eintrag im IKS-Blog).

Weiters ist das Team des SNML-TNG in der Rolle des Referenten bei mehreren Vorträgen vertreten:

• Sebastian Schaffert stellt das Paper „The Linked Media Framework: Integrating and Interlinking Enterprise Media Content and Data“ im Scientific Track des I-SEMANTICS vor (für die Zeit- und Ortsangaben des Vortrags verweisen wir auf das Programm).
• Dietmar Glachs stellt im Demo-Track der I-SEMANTICS unter dem Titel „Interlinking Media Archives with the Web of Data“ den Prototypen der Multimedia Annotation für das ORF Archiv vor (für die Zeit- und Ortsangaben des Vortrags verweisen wir auf das Programm).
• Georg Güntner stellt im I-PRAXIS-Track am 06.09.2012 um 14.45 Uhr Wege für den erfolgreichen Einsatz semantischer Technologien im Unternehmen vor. Der Titel des Referats lautet:  „Smart Enterprises: Successful implementation of Semantic Technologies in Enterprises„

Wir freuen uns auf ein Wiedersehen auf der I-SEMANTICS/I-KNOW 2012 in Graz. Besuchen Sie uns am Stand der Salzburg Research!

Below you find the text of the announcement on the mailing lists (by Sebastian Schaffert, 06.07.2012):

The Linked Media Framework (LMF) is a Linked Data Server with advanced functionalities like transparent Linked Data caching, querying, reasoning and versioning and has now been extended to a complete Linked Data Publishing suite that you can use for a number of frequent tasks:

• Linked Data Publishing of RDF or legacy data
• Semantic Search over your own data
• SKOS Thesaurus editing and content interlinking

For this purpose, the LMF Standalone Installation offers a combined and pre-configured installer for LMF Core, Google Refine and Apache Stanbol that simplifies the interaction between these components. Publishing Linked Data is made very easy with our framework.

With its built-in data management tools, we consider the LMF particularly useful for „casual“ Linked Data users who want to publish small to medium-sized datasets (up to a few hundred million triples), e.g. in the Open Government or Linked Enterprise area. Another common use case is building small Linked Data applications combining local data with Web data.

Other scenarios (e.g. with custom legacy->RDF converters) can easily be implemented on top of the LMF by building custom projects that build upon the LMF modules. See the documentation on the website to learn how this can be achieved.

List of Major Changes since 2.0

• Apache Stanbol integration for content enhancement and interlinking
• Google Refine integration for data management and Linked Data publishing
• Semantic Search UI Editor for quickly building Semantic Search Interfaces
• SKOSjs Editor to build SKOS thesauruses (uses SPARQL Query/Update)
• improved user and security management
• native SPARQL implementation mapping queries directly to SQL
• improved performance for big datasets (imports, querying)
• integrated book demo to show Semantic Search configuration

LMF Demos (including some who show the „media“ aspect) can be found at  http://labs.newmedialab.at. Note that this server is still running older versions of the LMF. We will upgrade it to the latest release in the coming week.

I hope you will find our software useful and can spare 10 minutes to try it out!

I’d like to conclude with a testimonial from François-Paul Servant, Renault, whom we presented a pre-release two weeks ago:

„You cannot imagine how happy I’ve been to see such a great and neat demo. You know, I’ve been a long time believer in SW, but I was beginning to give up hope on getting well polished tools – useful things that work and look good. This demo changed my mind. Thanks! I can’t wait to have LMF installed on my machine and use it!“