Using CBR to Estimate Development Effort for Web Hypermedia Applications

May 2002

Proceedings of the Fifteenth International Florida Artificial Intelligence Research Society Conference (FLAIRS 2002)

Volume

Proceedings of the Fifteenth International Florida Artificial Intelligence Research Society Conference (FLAIRS 2002)

All Papers

Good estimates of development effort play an important role in the successful management of larger software development projects. This paper compares the prediction accuracy of three CBR techniques to estimate the effort to develop Web hypermedia applications. Most comparative studies have used one CBR technique. We believe this may bias the results, as there are several CBR techniques that may also be used for effort predictionThis paper shows that a weighted Euclidian similarity measure was the most accurate of the CBR techniques tested.