G. Grieser and Y. Tanaka (Eds.): Intuitive Human Interface 2004, LNAI 3359, pp. 189200, 2004.  Springer-Verlag Berlin Heidelberg 2004 On Information Organization in Annotation Systems Panos Constantopoulos, Martin Doerr, Maria Theodoridou, and Manos Tzobanakis Institute of Computer Science, FORTH Heraklion, Crete, Greece {panos,martin,maria,tzoban}@ics.forth.gr Abstract. A rich semantic model of annotation is presented that distinguishes several annotation types, uses controlled vocabulary where appropriate, and supports multi-facetted characterization of documents, annotation of parts of documents as well as of entire documents, and annotation of document associa- tions besides annotation of single documents. This model has served as the ba- sis for implementing a powerful annotation system included in an advanced digital library system. 1 Introduction According to the New Oxford English Dictionary, an annotation is a note by way of explanation or comment added to a text or diagram. As such, annotations are primar- ily personal records of reading and interpretation, but they also often serve as shared records of work and opinion within specific scientific or professional communities. In the world of paper documents annotations are usually found either as arbitrary textual comments, or as characterizations according to specific aspects of interest. These can be unstructured and very personal, or they can be relatively structured, following certain rules with regard to aspects of annotation and format (usually established within a community). In what concerns the physical carrier of an annotation, this can be the annotated document itself or a separate document, usually a note card. In either case an annotation can be thought of as a distinct, yet secondary document referring to a primary one. In the world of digital documents, annotations are mostly encountered in the con- text of digital libraries, scientific databases and document management systems, and they share the purpose and nature of traditional paper annotations. They can be incor- porated in the documents, or they can be treated as (secondary) documents them- selves, referring to other documents, which can be documents of independent origin (primary documents), or other annotations. When treated as documents, annotations may reside in an autonomous annotation base, to which all the usual document man- agement services apply. Alternatively, they can be collocated with the documents they annotate. Digital annotations present higher potential than paper annotations for supporting discourse within scientific or professional communities, especially by virtue of the possibility offered by the World Wide Web to uniformly access distinct annotation servers. The cost to be paid for realizing this potential is a compelling need to ensure common interpretations of the data elements contained in the annotations both by the programs accessing and manipulating them and by the people using them. This, in turn, implies a need for common information organization schemes.