Many projects have to manage a vast collation of information, from Web pages, mailing lists, project pages (such as SourceForge), project documentation, etc. There is even sometimes valuable resource in such overlooked areas as IRC logs and meeting notes. Very often this information is poorly integrated, and more importantly, it is hard for someone who ploughs through a cross-section of it to easily document the thread of resources that were most useful to them in a particular need. The 4Suite project, developing an open source XML processing platform written in Python and C had such a problem. The author set about to solve this problem by developing a sort of graph of interconnected information and documentation snippets using RDF for extensible relationships between these data snippets and key metadata.

The heart of the Akara is an RDF model. Some of the RDF model nodes are the URIs of hosted documents. Documents in several formats are supported, but a rich, Wiki-like text (Akara markup) is the usual default information entry format, so that users can enter marked-up information rapidly. This data is presented using a series of XML Stylesheet Transform Language (EXSLT) stylesheets with embedded RDF queries for extracting key metadata for each information node. The engine includes a Wiki text to XML translator which parses Wiki text and emits a SAX 2 stream of XHTML for further processing. The RDF Query is embedded in XSLT using the Versa query language ([VERSA]), which is designed with XSLT users in mind. The whole system works on a Representational State Transfer (REST) idiom of simple read and modify query driven by URLs using Web forms. Among the modifications possible are the re-organization of information nodes after initial entry, based on addition and alteration of RDF metadata. This gives the system some of the qualities of a Wiki, but reduces the labor intensiveness of information reorganization. Information nodes and updates can be annotated and commented upon by other users in a manner remeniscent of blogs. This presentation covers examples and details of all these processing steps.

The Akara is an open knowledge management tool that emphasizes casual information entry and later collation and organization. More importantly, it uses generic XML-related technologies for this purpose. This presentation provides the detail for attendees to undertake similar projects either for general use, or in particular organizatonal situations.