ORES-2010 Ontology Repositories and Editors for the Semantic Web

Abstract

pCaoppeyrrsightby© 2th0e10 fpoarpethrse' inaduivthidoursa.l Copying permitted only for private and aepdcuaibtdloisershm.eidc paunrdposecos.pyTrihgihstedvolubmye itiss

-

10-Jun-2010: submitted by Christoph Lange 11-Jun-2010: published on CEUR-WS.org

Collaborative Metadata Editor Integrated with

Ontology Services and Faceted Portals

Jussi Kurki and Eero Hyv¨onen

Semantic Computing Research Group (SeCo) Aalto University, School of Science and Technology, and University of Helsinki http://www.seco.tkk.fi/, [e-mail] 1

Introduction

The basic process of implementing a faceted (semantic) portal [ 1–3 ] includes the following major steps: 1) Formulate vocabularies/ontologies/facets for representing domain concepts. 2) Design metadata schemas for representing content using (1). 3) Annotate content using (1) and (2), typically by a group of distributed peers in a Web 2.0 fashion. 4) Select the facets (1) and create the portal with semantic search and browsing facilities. This paper presents the tool SAHA 31 for the latter two phases. The idea is that given a set of domain vocabularies/ontologies/facets and a metadata schema for annotation (phases 1 and 2), a web-based annotation facility for distributed semantic content creation can be created instantly without programming skills. In a similar way, a faceted portal is automatically and instantly created online after annotating the content by just selecting facets for searching. Again, programming is not needed but only configuring the system using a web-based interactive interface.

We first describe the main features of the SAHA 3 metadata editor for the content creation phase, and then the integrated faceted portal engine HAKO2 for content publishing, followed by notes about implementation and scalability. Finally, contributions of the work w.r.t related systems are discussed, and some application use cases are listed.

1 http://www.seco.tkk.fi/services/saha/ 2 http://www.seco.tkk.fi/tools/hako/

SAHA 3 Metadata Editor Features

The original requirements for the web-based annotation editor SAHA [ 4 ] are simplicity (hiding technical concepts related to markup languages and ontologies from its user), adaptivity (to different metadata models), quality (helping and guiding the annotator to good and correct annotations), collaboration (supporting distributed simultaneous annotation at different locations), and portability (using the system on the web without installing any special software). SAHA 3 is a completely re-written version of SAHA with the following main new features: First, there is more support for general RDF editing, e.g. for inline editing of nested metadata, and for extending internal vocabularies. Second, the system is scalable to large datasets up to hundreds of thousands of objects. Third, SAHA 3 incorporates a simple publishing platform for building end-user search portals with full-text and multi-faceted search. The whole pipeline3 from metadata editing to the end-user portal application is accessible and configurable through a web-based interface.

During annotation, references to external ontologies are handled using the ONKI web service interface [ 5, 6 ]. SAHA 3 utilizes autocompletion [ 7, 8 ] as a key component to find references. When the user tries to find a concept, SAHA 3 uses at the same time web services to fetch concepts from connected external ONKI ontology repositories, and a local index to find locally defined concepts. Results are shown in one autocompletion result list regardless of origin. The same query can cover several ontology repositories at the same time. References to resources within the project at hand and external to it (in an external ontology repository) are transparent to the user.

The inline editor is new feature in SAHA 3 that has been found very handy by end-users. The idea is simple: a resource referenced through an object property can be edited inline at the right location in a small version of the editor inside the existing editor. In this way, several levels of editors can be opened recursively within each other, and the RDF network can be edited without moving from one resource window to another. For example, in Figure 1 the FOAF profile of the first author of this paper is edited, and the value of the property ”knows” is opened inline as a similar profile editor for the second author of this paper. Although the user interface easily becomes cluttered after a few levels, the inline editor is a handy way to add and edit nested metadata, as in this example shows.

SAHA 3 supports collaborative simultaneous editing. Resources that are being edited by one user are locked from other users. A chat facility has been implemented in the editor to facilitate instant discussions between peer editors (cf. the upper right corner in Fig. 1). 3

Faceted Search Engine HAKO

A SAHA 3 project can be published through the search interface of the HAKO portal engine. HAKO supports both free-text and faceted search [ 1–3 ]. For exam3 SAHA Sandbox: http://demo.seco.tkk.fi/saha3sandbox/saha3/main.shtml ple, Fig. 2 depicts a HAKO application using the RDF store of the Kirjasamposystem4 containing tens of thousands of instances of literary work of different kinds. The user has entered the keyword ”tolstoi”. Free-text search is done as a prefix search by default. On the left, one can see the hit list of 70 books and plays related to Tolstoi distributed over the facet categories, from where the user can refine the search either by type or by theme. The facets are configurable—the administrator can select any object property to be a facet top category.

Apart from the facet administration, HAKO interface is actually only a frontend to the SAHA 3 model. The configurable, shared data model (including the indices), and the dataset used by SAHA 3 are the same as those used by HAKO. This means that all modifications made with the SAHA 3 editor are reflected on the HAKO interface instantly.

The system is implemented in Java on top of Spring5 framework. The data model is based on TDB6 RDF database. Full-text search is backed by Lucene7. The editor interface is built using DWR8 and Dojo9 AJAX-components.

4 http://kirjasampo.fi/

5 http://www.springsource.com/ 6 http://openjena.org/TDB/ 7 http://lucene.apache.org/ 8 http://directwebremoting.org/ 9 http://www.dojotoolkit.org/ SAHA 3 makes use of external distributed ONKI ontology repository services, which is very handy when utilizing external third party ontologies and large vocabularies. From the interface design viewpoint, inline editing has been found a very useful feature. A major novelty of the system is the integration of the editor with a portal engine in real time. The idea of instant creation of a faceted search engine bears resemblance with SIMILE Exhibit10. However, in our case the system is integrated with a metadata editor, ontology services, and Semantic Web data models, and the search technology is based on Lucene (and not JavaScript) scaling up to very large datasets. SAHA 3 has been used in projects containing over 100,000 instances, and it has not shown signs of slowing down, or requiring large amounts of memory. The Tomcat server11 used runs well below 500MB with large projects.

SAHA 3 is in production use in the HealthFinland12 system and the Kirjasampo project, where some 50 librarians have been annotating metadata about tens of thousands of novels, short stories, authors, and other objects related to Finnish literature for a semantic portal. The editor has been used is many ways 10 http://www.simile-widgets.org/exhibit/ 11 http://tomcat.apache.org/ 12 http://www.seco.tkk.fi/applications/tervesuomi/ in CultureSampo13, e.g. for creating the semantic narrative descriptions of the Finnish History Ontology14 and the Semantic Kalevala epic15. HAKO has been used in several industrial application demonstrations for searching documents. Acknowledgements This work is part of the National Semantic Web Ontology project in Finland16 (FinnONTO, 2003–2012), funded mainly by the National Technology and Innovation Agency (Tekes) and a consortium of 38 organizations. 13 http://www.seco.tkk.fi/applications/kulttuurisampo/ 14 http://www.seco.tkk.fi/ontologies/histo/ 15 http://www.seco.tkk.fi/applications/kulttuurisampo/kalevala/ 16 http://www.seco.tkk.fi/projects/finnonto/

References

1. Pollitt , A.S.: The key role of classification and indexing in viewbased searching . Technical report , University of Huddersfield, UK ( 1998 ) http://www.ifla.org/IV/ifla63/63polst.pdf.

2. Hearst , M. , Elliott , A. , English , J. , Sinha , R. , Swearingen , K. , Lee , K.P. : Finding the flow in web site search . CACM 45 ( 9 ) ( 2002 ) 42 - 49

3. Hyv¨onen, E., M¨akela¨, E., Salminen , M. , Valo , A. , Viljanen , K. , Saarela , S. , Junnila , M. , Kettula , S.: MuseumFinland-Finnish museums on the semantic web . Journal of Web Semantics 3 ( 2 ) ( 2005 )

4. Valkeap¨aa¨, O. , Alm , O. , Hyv¨onen, E.: A framework for ontology-based adaptable content creation on the semantic web . Journal of Universal Computer Science 13 ( 12 ) ( 2007 )

5. Viljanen , K. , Tuominen , J. , Hyv¨onen, E.: Ontology libraries for production use: The Finnish ontology library service ONKI . In: Proceedings of the ESWC 2009 , Heraklion, Greece, Springer-Verlag ( 2009 )

6. Tuominen , J. , Frosterus , M. , Viljanen , K. , Hyv¨onen, E.: ONKI SKOS server for publishing and utilizing SKOS vocabularies and ontologies as services . In: Proceedings of the ESWC 2009 , Heraklion, Greece, Springer-Verlag ( 2009 )

7. Hyv¨onen, E., M¨akel¨a, E.: Semantic autocompletion . In: Proceedings of the First Asia Semantic Web Conference (ASWC 2006 ), Beijing, Springer-Verlag ( 2006 )

8. Hildebrand , M., van Ossenbruggen , J. , Amin , A. , Aroyo , L. , Wielemaker , J. , Hardman , L. : The design space of a configurable autocompletion component . Technical Report INS-E0708 , Centrum voor Wiskunde en Informatica , Amsterdam ( 2007 )