PRISM (PRogramming In Statistical Modeling) is a logic-based language that integrates logic programming, probabilistic reasoning, and EM learning. It allows for the description of independent probabilistic choices and their consequences in general logic programs. PRISM supports parameter learning. For a given set of (possibly incomplete) observed data, PRISM can estimate the probability distributions that best explain the data. This power is suitable for applications that include 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 when compared with specific statistical tools, such as Hidden Markov Models (HMMs), Probabilistic Context Free Grammars (PCFGs), and discrete Bayesian networks. Thanks to the good efficiency of the linear tabling system in B-Prolog, and thanks to the EM learner adopted in PRISM, PRISM is comparable in performance to specific statistical tools on relatively large amounts of data. PRISM is a product of the PRISM team that is led by Taisuke Sato at the Tokyo Institute of Technology.
Neng-Fa Zhou
2013-01-25
