Loopy Propagation in a Probabilistic Description Logic
SUM '08 Proceedings of the 2nd international conference on Scalable Uncertainty Management
| Hi-index | 0.00 |
The work described in this paper was inspired by a problemincreasingly vexatious to many businesses confronting theever-diminishing life cycles of modern productsviz., that ofpredicting characteristics (such as overall demand, segmentation,etc.) of markets facing new product introductions. A framework isproposed that allows the market parameters of new products to bederived by analogy with those of old ones. To do so, the frameworkcombines the capabilities of Bayesian networks [14] anddescription logics [16]. The paper commences with anexposition of the problem that motivates the work. There followbrief descriptions of Bayesian networks and description logics intheir unalloyed state, and a discussion of issues surrounding theircombination. The combined system is presented, along with anaccount of its formal details and an inference procedure forentailment. A sample application of the framework is given. Thepaper concludes by comparing the proposed framework with relatedexisting systems and by suggesting possible courses for futuredevelopment.