Lean knowledge management is today implemented mostly through wikis, which let users enter text and

other data, such as files, and connect the content through hyperlinks. Easy setup and a huge variety of

editing support are primary reasons for wiki use in all types of intranet- and Internet-based information

sharing (see P. Louridas, “Using Wikis in Software Development,” IEEE Software, Mar. 2006, pp. 88–

91). The drawbacks show up when you need to structure data as opposed to just edit text. Many wikis

have tons of useful content, but the volume and lack of structure make it inaccessible over time. This is

where semantic wikis enter the picture. Sebastian Schaffert and his colleagues describe them here and explain

how to model wiki knowledge and content for improved usability. I look forward to hearing from both readers and prospective authors about this column and the technologies you want to know more about. —Christ of Ebert

semantic wikis combine wiki properties, such as ease of use, collaboration, and linking, with Semantic Web technologies, such as structured content, knowledge models in the form of ontologies, and reasoning. Semantic wikis connect social and artificial intelligence, supporting users in ways that aren’t available in normal wikis. They make it easy to introduce novel Semantic Web technologies that are arguably still hard to grasp in other applications. Here, we give an overview of semantic wikis, focusing on application fields and briefly comparing selected systems on the basis of their different approaches.

Wikis and semantic wikis

A wiki is a Web-based system that enables collaborative editing of Web pages. The most important properties of wikis are their openness and flexibility. Their openness lets each user participate in content creation, and their flexibility supports different users’ working styles without imposing technological constraints. Wikis provide a

Web-based text editor with a simple markup language to create content and to link easily between pages as well as a versioning system to track content changes and full-text search for querying the wiki pages.

However, full-text search often proves insufficient for retrieving knowledge such as structured data or related pages. For instance, consider Wikipedia. It contains abundant structured data related to persons (date and place of birth, important works, and so on) and cities (geographical coordinates, relations such as located-in, and so on). Wikis generally have two ways to make such data more accessible. On one hand, most wikis contain manually updated overview pages that sort pages according to certain criteria. This approach usually involves a lot of maintenance work. On the other hand, most wiki systems offer additional tools such as categories, extensions for certain metadata

types (for example, calendars), and template mechanisms that predefine the structure for certain kinds of pages. However, the category systems tend to be inflexible, specific extensions are restricted, and their benefits don’t necessarily outweigh the time it takes users to learn them. A semantic wiki tries to extend a normal wiki’s flexibility to address structured data. To this end, it supports metadata in the form of semantic annotations of the wiki pages themselves and of the link relations between wiki pages. The annotations usually correspond to an ontology the fines the properties that can be associated with different object types. Wiki users edit and maintain the ontology within

the semantic-wiki system, using knowledge models that are usually represented in RDF Schema and OWL (Web Ontology Language). Once the semantic annotations are created, they are then available for extendedqueries or even for adaptation of the content presentation to different users and domains. The term “semantic wiki” applies to a

variety of different systems. The internal representation of annotations with RDF/ OWL simplifies the data exchange with other applications. For example, it supports external-search functionality, such as a Web service. Furthermore, the wiki system can use deductive reasoning to derive additional information for the user. In addition, semantic wikis offer a simple formalism for semantically annotating links and wiki articles or other kinds of content; a semantic search for querying by not only keyword but also semantic relations between objects—for example, by topic (“EU projects”) or even indirectly\ (“meeting minutes of EU projects”); and possibly an additional automatic or semiautomatic extraction of metadata from wiki articles to simplify the annotation process. Different semantic-wiki systems follow different goals. Some aim to simplify navigation and collaboration by using semantic annotations, while others want to establish wikis as a means to collaboratively create Semantic Web ontologies.

The salient research-and-development aspect of semantic wikis thus lies between the adoption of Semantic Web technologies for wikis and the use of wikis for the Semantic Web. In a sense, a semantic wiki can therefore be considered “the Semantic Web in small.”

Semantic-wiki example

Semantic-wiki users create and maintain the ontology in a semantic wiki by associating each ontology instance and concept with a wiki page. Wiki links and annotations relate the concepts and pages with each other. Semantic wikis follow two annotation approaches: while most provide an extended wiki link syntax within the textual editor,

some provide form- or Ajax-based annotations in a separate annotations editor. Figure 1 shows the textual annotation

in Semantic MediaWiki.1 For instance, the annotation [[coded in::Java]] denotes a “coded” relationship between the concepts Ike Wiki and Java. This formal representation is now available for querying all semantic wikis that are written in Java. Figure 2 shows form-based annotation in the IkeWiki system.2 The annotation is for the term “Bilberry” in a biology application. The system lets users create and annotate not only links and pages but also

ontology classes and properties. Immediately below the page title are the associated article types—in the example, bio: Species,  skos  :Concept, and  rdfs: Resource. Icons in the text mark already-annotated links. Relations to other wiki articles and ontology concepts are listed under the References box on the right side. This list visualizes the existing annotations. The references also aim to ease navigation through the evolving semantic network. Clicking on the “+” symbols associated with a link or type opens a form that lets you add annotations to the page. Figure 2 demonstrates the annotation of a link between the subject Bilberry and the object Vaccni um with the relation bio:isSpeciesFor. The user can choose from a list of properties defined in an underlying knowledge model and doesn’t need a priori knowledge about the available annotations.

System comparisons

As with normal wikis, there is no “standard” semantic wiki. Each system has its own focus, strengths, and weaknesses. The following summaries of five semantic-wiki systems come with a caveat: most of the

systems are currently available only as research prototypes, so their features change quickly.

Semantic Media Wiki

This semantic wiki concentrates on the Wikipedia-encyclopedia scenario and thereforeemphasizes scalability and backward compatibility.1 No predefined schema or ontology is required for annotations, so users can add new annotations as needed—similar to tagging systems. Because efficient and freely available inference systems that scale

up to the size of Wikipedia aren’t foreseeable in the near future, Semantic MediaWiki doesn’t support inferencing and similar advanced functionalities. Its Web site is http:// ontoworld.org/wiki/Semantic_MediaWiki.

Ike Wiki

This Java Web application was originally developed as a tool for creating ontologies collaboratively and for managing knowledge. 2 To support users in these tasks, Ike Wiki focuses primarily on providing advanced semantic functionalities such as reasoning. This different focus allows Ike Wiki to accept lower scalability and

higher hardware demands than Semantic Media Wiki. Ike Wiki supports developers in using as well as editing OWL ontologies. You can configure it to use OWL-RDF Schema or OWL DL (description logics) reasoning. A

rule-based inference mechanism is under development. Ike Wiki’s Web site is http://ikewiki. salzburgresearch.at.

Kaukolu

This research prototype is based on JSP Wiki.  Kaukolu allows annotations with extended wiki markup as well as form-based annotations that are built dynamically from underlying ontologies. Annotations can refer to arbitrary parts of a page rather than just the whole page, and external systems can generate annotations automatically. For example, experiments are currently under way that use eyetracking technology and an eyetracker-based extension to highlight

text. One application scenario is the annotation of existing documents such as juridical texts. The Web site is http://kaukoluwiki. opendfki.de.

SweetWiki

This research prototype from Inria Sophia- Antipolis is implemented in Java. Sweet- Wiki combines social tagging with formal ontologies. Users can easily annotate pages with arbitrary tags, which they can in turn associate with concepts from the underlying ontologies. In addition,  Sweet Wiki uses the Corese inference machine, which was developed for

conceptual graphs and offers many reasoning services. The Web site is http://www-sop inria.fr/acacia/soft/sweetwiki.html.

Onto Wiki

This system differs from the others we’ve mentioned in that classical textual content is no longer in the foreground. Instead, Onto Wiki offers an easy-to-use interface for collaboratively creating and maintaining ontologies.3 It also supports semantic search and navigation as well as the possibility of versioning metadata. The Onto Wiki Web site is http://3ba.se. It includes an extensive overview of semantic wikis (http://wiki.ontoworld.org/wiki/ Category: Semantic _wiki).

Application areas

As an extension of normal wikis, semantic wikis obviously overlap a large area of applications with them. However, by using explicit metadata representations, semantic wikis offer significant support for improved navigation and search, context dependent presentation, personalization, and other such features. Here, we present two example application areas that illustrate the different aspects of using semantic wikis.

Knowledge management

Recently, wikis are increasingly used as tools to support knowledge management. For example, many companies use wikis to maintain and share knowledge about software projects (source code, documentation, project work plans, bug reports, and so on). Knowledge captured in this fashion is easy to create but increasingly difficult to retrieve. It’s often even distributed over a multitude of different wiki installations. For example, more than a dozen wikis currently support Sun Microsystems’ Net Beans integrated development environment, but none of them is integrated with

the others. Semantic wikis have the potential to solve these problems without sacrificing wiki flexibility and openness. Where semantic annotations and hence structures are available, the system can actively support the user by—for example—appropriately visualizing semantically represented project plans, exchanging annotations with other wikis or even applications, or offering a semantic-search function. Where semantic annotations aren’t yet available, a semantic wiki still offers the same functionality as an ordinary wiki. Like in an ordinary wiki, where many short-term users often contribute content but a few long-term users (“gardeners”) constantly maintain the system, long-term

users in semantic wikis take care of properly structuring the content. A nice side effect

of this “evolutionary formalization” is that users can instantly benefit from the additional effort of annotation, which is a crucial factor for motivating users.

Several European Union projects are investigating knowledge management in semantic wikis. These include KiWi (Knowledge in a Wiki, www.kiwi-project.eu) and Nepomuk  (Networked Environment for Personalized,

Ontology-Based Management of Unified Knowledge, http://nepomuk. semanticdesktop.org).

Ontology engineering

Developing a Semantic Web ontology is daunting. Domain experts, such as biologists, know their domain well but lack expertise in knowledge management formalisms. Computer scientists, on the other hand, have sufficient knowledge about the formalisms but not about the respective domains. For this reason, most high-quality ontologies outside computer science exist in special areas such as medicine or biology where the high effort is

acceptable. Semantic wikis can significantly simplify ontology engineering. Starting from textual descriptions in wiki pages created by domain experts, computer scientists can interact closely with the experts to successively

formalize knowledge. Other users can further refine the ontologies that evolve this way for use in other tools or

applications. Semantic wikis can also help update and maintain an ontology. In tools such as Protégé, domain experts need to understand the underlying formalism, although they can perform most changes in a semantic wiki. Early experiences with the EU Dynamont project are promising (http://dynamont.factlink.net).

by offering an explicit representation of the knowledge in wikis, semantic wikis denote a lively research area that considers many questions in the context of the Semantic Web initiative. For semantic wikis, system usability is of prime importance to ensure that untrained users can employ them in an ad hoc manner. Other research areas include the integration of semantic query technologies, efficient automated reasoning, semiauto mated annotation,

and user adaptability.

فيلسوفان اقدم يوناني در جهان پيرامون خويش دگرگوني مشاهده مي‌كردند – چرخه فصلها، زايش و رشد و مرگ موجودات زنده، بخار شدن آب به هنگام گرما و به جاي اول بازگشتن آن از هوا – و به اين پرسش برانگيخته مي‌شدند كه چگونه بايد درباره اين تغييرات سخن بگويند و سبب اين امور را چطور بيان كنند. در اساطير ديرين به همين بسنده مي‌شد كه بگويند اراده متلون و بلهوسانه برخي موجودات شبيه به انسان و نامقيد به قوانين عام طبيعت در اين كار دخالت دارد. (يعني ايزدان اساطيري)

 فلسفه حوزه‌اي از دانش بشري است كه به پرسش و پاسخ درباره مسائل بسيار كلي و جايگاه انسان در آن مي‌پردازد؛ مثلاً اين كه آيا جهان و تركيب و فرآيندهاي آن به طور كامل مادي است؟ آيا به وجود آمدن يا به وجود آوردن جهان داراي هدف است؟ آيا ما مي‌توانيم پاسخ قطعي بعضي چيزها را بيابيم؟ آيا ما آزاد هستيم؟ آيا ارزشهاي مطلقي وجود دارند؟ تفاوت اصلي فلسفه يا علم در اين است كه پاسخهاي فلسفي را نمي‌توان با تجربه يا آزمايش تاييد كرد.