Background

Motivation

The development of appropriate technologies to enable direct and efficient access to relevant information is one of the key challenges for the knowledge society. Traditional database query languages are optimised to reconcile large volume data stocks with great efficiency based on exactly specified queries. Today, many application scenarios demand more advanced technologies that offer intelligent support to the user when searching for information. In e-Commerce, for example, the user searching for a product is seldom in a position to formulate an appropriate, exact specification for a query. Frequently, either the necessary background knowledge about the product offering is lacking or, over or under specified database queries lead to empty or unmanageable results, which is of little help in finding a suitable available product. Intelligent product recommendation systems present an alternative which, even with relatively vague desires and needs on the part of the user, can recommend suitable, target oriented products.

Such knowledge-based product recommendation systems are being developed primarily on the basis of technologies from the field of Case-Based Reasoning (CBR). The central idea here is to search on the basis of similarities to the stated enquiry, which always enables the return of the most appropriate, available information sorted by relevance/utility. A major role is played by the inclusion of knowledge domains in the form of ontologies and other application specific similarity measures.

myCBR Workbench and SDK

The development of even a quite simple CBR application already involves a number of steps, such as collecting case and background knowledge, modelling a suitable case representation, defining an accurate similarity measure, implementing retrieval functionality, and implementing user interfaces. Compared to other AI approaches, CBR allows to reduce the effort required for knowledge acquisition and representation significantly, which is certainly one of the major reasons for the commercial success of CBR applications. Nevertheless, implementing a CBR application from scratch remains a time consuming software engineering process and requires a lot of specific experience beyond pure programming skills.

Although CBR research has a history of over 20 years, and in spite of the broad commercial success of CBR applications in recent years, today only few CBR software tools for supporting the development process are available. The key motivation for implementing myCBR was the need for a compact and easy-to-use tool for building prototype CBR applications in teaching, research, and small industrial projects with low effort. Moreover, the tool should be easily extensible in order to facilitate the experimental evaluation of novel algorithms and research results. Therefore, it provides comfortable graphical user interfaces for modelling various kinds of attribute-specific similarity measures and for evaluating the resulting retrieval quality. In order to reduce also the effort of the preceding step of defining an appropriate case representation, it includes tools for generating the case representation automatically from existing raw data. The Software Development Kit (SDK) allows for easy integration into other applications and extension to specific requirements such as additional similarity calculations.

Research background

myCBR's foundation can be retraced to several European research projects, especially INRECA, INRECA II, and WEBSELL:

© 2006-2015 DFKI GmbH | Legal Information | | Logo Design by Stefan Schulz
valid XHTML 1.1 and CSS 2.1