Empolis orenge - an Open Platform for Knowledge Management Applications
暂无分享,去创建一个
empolis orenge 4.0, a component-based environment for knowledge management applications by empolis Knowledge Management, is a flexible and scalable Case-Based Reasoning shell for industrial applications, that also contains many components that provide functionality beyond the basic CBR paradigm. Further it can be extended easily by application-specific components, which makes it an ideal starting point for experimental CBR-oriented research projects, too. As Case-Based Reasoning has evolved into a well-established problem solving paradigm, industrial CBR applications are getting more complex and larger. Moreover, the typical CBR tasks like Retrieval are usually only a small part of the required functionality. Often the bigger challenge is to integrate the CBR tool neatly into an existing information system environment and to combine CBR with other techniques for a better solution to the problem. Also, as CBR is used for enterprise-wide knowledge management systems or for product search in huge electronic malls, the CBR tool must be able to cope with huge industrial case bases containing a million or more cases. From this, three major requirements emerge that a CBR tool for the 21st century must meet: Open Architecture It must be easy to combine components implementing different techniques or to enhance the tool by components implementing new techniques or application specific tasks. Connectivity It must be easy to use existing data in the application environment and to integrate the tool in existing information systems. Scalability It must be possible to handle huge amounts of data efficiently. empolis orenge has been developed by empolis Knowledge Management (formerly tec:inno), and has a lot of experience with developing and deploying CBR projects, also using the predecessor products of orenge, CBR-Works (see [3]), CBR-Sells, and CBR-Answers (for both see [1]). Based on this experience, empolis orenge has been designed to meet the requirements stated above: – orenge is component-based: Instead of being a big monolithic program, orenge consists of a set of components or services, that perform only one special task.
[1] Dietmar Janetzko,et al. Case Retrieval Nets as a Model for Building Flexible Information Systems , 2001, Künstliche Intell..
[2] Ralph Bergmann,et al. An Efficient Approach to Similarity-Based Retrieval on Top of Relational Databases , 2000, EWCBR.
[3] Stefan Schulz. CBR-Works A State-of-the-Art Shell for Case-Based Application Building , 1999 .