SemDAV
Today, the organization of and human interaction with digital data assets is heavily influenced by hierarchical file systems. However, steadily increasing volumes of data require new paradigms for data organization. In order to support people who work with digital information, these approaches should more closely parallel how the human brain stores and retrieves information.
The goal of the SemDAV project is to research basic technologies that enable the creation of systems for semantically-enabled information management. SemDAV will extend the well-known WebDAV protocol and provides a platform- and implementation-independent method for exchanging content and associated semantic metadata. Instead of storing mere information (i.e. simple attributes) about digital objects, systems based on this new protocol will be able to process descriptive knowledge about such objects. These semantic systems are able to handle complex typed attributes, multi-dimensional classification schemes, and in particular relationships between data, time, location, users, and applications - knowledge which places digital objects in an intelligent context.
At the same time the protocol is being specified, two prototypes will also be implemented: On the server side, a reference implementation will demonstrate the feasibility of our approach. Therein, we will implement intelligent methods for the extraction of metadata from various sources, including searching and browsing behaviour, content-based analysis, and manual annotations by users and applications. On the client side, we will develop a user interface which demonstrates efficient methods of representing the collected content and metadata, and assists the user in navigating and searching the data inventory more efficiently and intuitively.
In a long-term perspective, products based on SemDAV may replace storage systems for unstructured content, such as shared file servers or personal home directories. SemDAV will explore new ways for users and applications to interact with files, and may serve as a key enabling technology for intelligent data management.