Knowledge in a Wiki

In recent years, wikis have become a popular means for enabling collaborative work on emerging content, allowing users to work together to author, edit and administrate wiki pages. Content in traditional Wikis consists of natural language text (and possibly multimedia files) and is not directly accessible to automated semantic processing. Therefore, knowledge in Wikis can be located only through simple user-generated structures (tables of contents, inter-page links) and full text search. More advanced functionalities that are highly desirable in knowledge-intensive contexts like querying, reasoning and semantic browsing are not possible. Semantic Wikis introduce capabilities into Wikis for specifying knowledge not just in natural language but also in more formal, machine-processable ways.

         However, many of the technologies that Semantic Wikis employ were developed for use in a static environment with annotations and rules being crafted by knowledge representation experts. This is in contraposition to the ever-changing, dynamic character of Wikis where content and annotations are, for the most part, created by casual users. The EU-funded project "KiWi – Knowledge in a Wiki" is concerned with development of a next-generation Semantic Wiki with extended functionalities in the areas of personalization (University of Aalborg, Denmark), information extraction (University of Brno, Czech Republic) and, finally, reasoning and querying, the focus of the work of our KiWi team at the University of Munich.

         One of the main goals of the KiWi team at the University of Munich is the development and implementation of a user-friendly rule-based query language for the KiWi wiki that can combine queries over data, annotation and structure. In this regard, the keyword-based query language KWQL was developed together with its visual counterpart visKWQL and a system for approximate matching over structured data, PEST. KWQL allows for combined queries over full-text, annotations and content structure, fusing approaches from conventional query languages with information retrieval techniques for search. KWQL aims to make data contained in a semantic wiki accessible to all users—not only those who have experience with query languages. Queries have little syntactic overhead and aim at being only as complex as necessary. Further, KWQL has a flat learning curve and the complexity of queries increases with the complexity of the user's information need. Simple KWQL queries consist of a number of keywords and are no more complicated to write than search requests in web search engines. On the other hand, advanced KWQL queries can impose complex selection criteria and even reformat and aggregate the results into new wiki pages, giving rise to a simple form of reasoning. Our team also developed an inconsistency tolerant rule language for annotations, KWRL, and several reason maintenance based algorithms for incremental updates of materialized facts.