PRISM User ’ s Manual ( Version 2 . 0 beta 4 )

Preface The past several years have witnessed a tremendous interest in logic-based probabilistic learning as testified by the number of formalisms and systems and their applications. Logic-based probabilistic learning is a multidisciplinary research area that integrates relational or logic formalisms, probabilistic reasoning mechanisms, and machine learning and data mining principles. Logic-based probabilistic learning has found its way into many application areas including bioinformatics, diagnosis and troubleshooting, stochastic language processing, information retrieval, linkage analysis and discovery, robot control, and probabilistic constraint solving. PRISM (PRogramming In Statistical Modeling) is a logic-based language that integrates logic programming and probabilistic reasoning including parameter learning. It allows for the description of independent probabilistic choices and their consequences in general logic programs. PRISM supports parameter learning, i.e. for a given set of possibly incomplete observed data, PRISM can estimate the probability distributions to best explain the data. This power is suitable for applications such as learning parameters of stochastic grammars, training stochastic models for gene sequence analysis, game record analysis, user modeling, and obtaining probabilistic information for tuning systems performance. PRISM offers incomparable flexibility compared with specific statistical tools such as hidden Markov models (HMMs) [4, 30], probabilistic context free grammars (PCFGs) [4] and discrete Bayesian networks. PRISM employs a proof-theoretic approach to learning. It conducts learning in two phases: the first phase searches for all the explanations for the observed data, and the second phase estimates the probability distributions by using the EM algorithm. Learning from flat explanations can be exponential in both space and time. To speed up learning, the authors proposed learning from explanation graphs and using tabling to reduce redundancy in the construction of explanation graphs. The PRISM programming system is implemented on top of B-Prolog (http://www.probp.com/), a constraint logic programming system that provides an efficient tabling system called linear tabling [49]. Tabling shares the same idea as dynamic programming in that both approaches make full use of intermediate results of computations. Using tabling in constructing explanation graphs resembles using dynamic programming in the Baum-Welch algorithm for HMMs and the Inside-Outside algorithm for PCFGs. Thanks to the good efficiency of the tabling system and the EM learner adopted in PRISM, PRISM is comparable in performance to specific statistical tools on relatively large amounts of data. The theoretical side of PRISM is comprehensively described in [39]. For an implementational view, please refer to [50]. The user is assumed to be familiar with logic …