Introduction

Information systems often store data which describe real-world entities. In application projects it has been observed that multimedia content often has the very same function as data found in databases for, e.g., business applications. Viewed from this perspective, content management systems store views of entities that cannot fully be described by structured data alone.

Observations from Semiotics and Epistemology inspired the idea of Concept-oriented Content Management (COCoMa), which regards indivisible pairs of multimedia content and a conceptual model for its interpretation to describe entities in COCoMa systems.

Content is meant to represent information to system users. Since users work in certain contexts - their expertise, the task at hand, etc. - content has to be structured, managed, and presented with regard to individual information needs. This raises the demand for openness and dynamics of COCoMa systems. Users have to be able to model entities without having to regard technological constraints and they have to be able to keep their models open to changes. A COCoMa system reflecting a user's model has to adapt to changes dynamically.

To this end, COCoMa systems are designed to support information personalization. This includes both schema evolution and the management of variants of data. Users are able to refine schemata and data which are publicly available. COCoMa systems allow to access both public and personal models and data through an integrated view. Finally, when users have produced results they want to communicate to others, they can publish their findings, thus integrating their information with that publicly available.

To achieve openness and dynamics, the Conceptual Content Management approach consists of three major contributions:

1. a conceptual modeling language for domain modeling,
2. a model compiler which translates domain models to COCoMa systems, and
3. a software architecture designed to support systems evolution.