TBox Reasoning in the Probabilistic Description Logic SHIQ P

Abstract. One shortcoming of classic Descriptions Logics, DLs, is their inability to encode probabilistic knowledge and reason over it. This is, however, a strong demand of some modern applications, e.g. in biology and healthcare. Therefore, probabilistic extensions of DLs are attracting attention nowadays. We introduce the probabilistic DL SHIQP which extends a known probabilistic DL. We investigate two reasoning problems for TBoxes: deciding consistency and computing tight probability bounds. It turns out that both problems are not harder than reasoning in the classic counterpart SHIQ. We gain insight into complexity sources