Giuseppe Cota

cplint on SWISH: Probabilistic Logical Inference with a Web Browser

cplint on SWISH is a web application that allows users to perform reasoning tasks on probabilistic logic programs. Both inference and learning systems can be performed: conditional probabilities with exact, rejection sampling and Metropolis-Hasting methods. Moreover, the system now allows hybrid programs, i.e., programs where some of the random variables are continuous. To perform inference on such programs likelihood weighting and particle filtering are used. cplint on SWISH is also able to sample goals’ arguments and to graph the results. This paper reports on advances and new features of cplint on SWISH, including the capability of drawing the binary decision diagrams created during the inference processes.

Learning Probabilistic Description Logics

We consider the problem of learning both the structure and the parameters of Probabilistic Description Logics under the DISPONTE semantics. DISPONTE is based on the distribution semantics for Probabilistic Logic Programming and assigns a probability to assertional and terminological axioms. The system EDGE, given a DISPONTE knowledge base KB and sets of positive and negative examples in the form of concept assertions, returns the value of the probabilities associated with axioms. We present the system LEAP that learns both the structure and the parameters of DISPONTE KBs explotiting EDGE. LEAP is based on the system CELOE for ontology engineering and exploits its search strategy in the space of possible axioms. LEAP uses the axioms returned by CELOE to build a KB so that the likelihood of the examples is maximized. We present experiments showing the potential of EDGE and LEAP.

A survey of lifted inference approaches for probabilistic logic programming under the distribution semantics

Lifted inference aims at answering queries from statistical relational models by reasoning on populations of individuals as a whole instead of considering each individual singularly. Since the initial proposal by David Poole in 2003, many lifted inference techniques have appeared, by lifting different algorithms or using approximation involving different kinds of models, including parfactor graphs and Markov Logic Networks. Very recently lifted inference was applied to Probabilistic Logic Programming (PLP) under the distribution semantics, with proposals such as L P 2 and Weighted First-Order Model Counting (WFOMC). Moreover, techniques for dealing with aggregation parfactors can be directly applied to PLP. In this paper we survey these approaches and present an experimental comparison on five models. The results show that WFOMC outperforms the other approaches, being able to exploit more symmetries. The paper presents a survey of lifted inference for probabilistic logic programming.Lifted inference aims at answering queries from statistical relational models.Reasoning is made on the whole set of individuals in the population instead of considering each individual singularly.To deal with PLP, noisy-OR gates must be handled.A experimental comparison of three techniques for lifted inference for PLP is presented.

Distributed Parameter Learning for Probabilistic Ontologies

Representing uncertainty in Description Logics has recently received an increasing attention because of its potential to model real world domains. EDGE, for “Em over bDds for description loGics paramEter learning”, is an algorithm for learning the parameters of probabilistic ontologies from data. However, the computational cost of this algorithm is significant since it may take hours to complete an execution. In this paper we present \(\mathrm {EDGE}^{\mathrm {MR}}\), a distributed version of EDGE that exploits the MapReduce strategy by means of the Message Passing Interface. Experiments on various domains show that \(\mathrm {EDGE}^{\mathrm {MR}}\) significantly reduces EDGE running time.

Scaling Structure Learning of Probabilistic Logic Programs by MapReduce

Probabilistic Logic Programming has been shown to be a useful language for Inductive Logic Programming: for instance, the system SLIPCOVER learns high quality theories in a variety of domains. However, the computational cost of SLIPCOVER is sometimes expensive, with a running time of the order of hours. In this paper we present the system SEMPRE for “Structure lEarning by MaPREduce”, that implements SLIPCOVER by applying a particularly simple MapReduce strategy, directly implemented with the Message Passing Interface. SEMPRE has been tested on various domains and shown to effectively reduce SLIPCOVER running time, even if the speedup is often sublinear.

Tableau reasoning for description logics and its extension to probabilities

The increasing popularity of the Semantic Web drove to a widespread adoption of Description Logics (DLs) for modeling real world domains. To help the diffusion of DLs, a large number of reasoning algorithms have been developed. Usually these algorithms are implemented in procedural languages such as Java or C++. Most of the reasoners exploit the tableau algorithm which features non-determinism, that is not easily handled by those languages. Prolog directly manages non-determinism, thus is a good candidate for dealing with the tableau’s non-deterministic expansion rules. We present TRILL, for “Tableau Reasoner for descrIption Logics in proLog”, that implements a tableau algorithm and is able to return explanations for queries and their corresponding probability, and TRILLP, for “TRILL powered by Pinpointing formulas”, which is able to compute a Boolean formula representing the set of explanations for a query. Reasoning on real world domains also requires the capability of managing probabilistic and uncertain information. We show how TRILL and TRILLP can be used to compute the probability of queries to knowledge bases following DISPONTE semantics. Experiments comparing these with other systems show the feasibility of the approach.

A web system for reasoning with probabilistic OWL

We present the web application Tableau Reasoner for descrIption Logics in proLog on SWI‐Prolog for SHaring (TRILL on SWISH) which allows the user to write probabilistic description logic (DL) theories and compute the probability of queries with just a web browser. Various probabilistic extensions of DLs have been proposed in the recent past, because uncertainty is a fundamental component of the Semantic Web. We consider probabilistic DL theories following our distribution semantics for probabilistic ontologies (DISPONTE) semantics. Axioms of a DISPONTE knowledge base can be annotated with a probability, and the probability of queries can be computed with inference algorithms. TRILL is a probabilistic reasoner for DISPONTE knowledge base that is implemented in Prolog and exploits its backtracking facilities for handling the non‐determinism of the tableau algorithm. TRILL on SWISH is based on SWISH, a recently proposed web framework for logic programming, based on various features and packages of SWI‐Prolog (e.g., a web server and a library for creating remote Prolog engines and posing queries to them). TRILL on SWISH also allows users to cooperate in writing a probabilistic DL theory. It is free, open, and accessible on the Web at the url: http://trill.lamping.unife.it; it includes a number of examples that cover a wide range of domains and provide interesting Probabilistic Semantic Web applications. By building a web‐based system, we allow users to experiment with probabilistic DLs without the need to install a complex software stack. In this way, we aim to reach out to a wider audience and popularize the Probabilistic Semantic Web. Copyright

Probabilistic Inductive Logic Programming on the Web

Probabilistic Inductive Logic Programming (PILP) is gaining attention for its capability of modeling complex domains containing uncertain relationships among entities. Among PILP systems, cplint provides inference and learning algorithms competitive with the state of the art. Besides parameter learning, cplint provides one of the few structure learning algorithms for PLP, SLIPCOVER. Moreover, an online version was recently developed, cplint on SWISH, that allows users to experiment with the system using just a web browser. In this demo we illustrate cplint on SWISH concentrating on structure learning with SLIPCOVER. cplint on SWISH also includes many examples and a step-by-step tutorial.

Probabilistic Logical Inference on the Web

cplint on SWISH is a web application for probabilistic logic programming. It allows users to perform inference and learning using just a web browser, with the computation performed on the server. In this paper we report on recent advances in the system, namely the inclusion of algorithms for computing conditional probabilities with exact, rejection sampling and Metropolis-Hasting methods. Moreover, the system now allows hybrid programs, i.e., programs where some of the random variables are continuous. To perform inference on such programs likelihood weighting is used that makes it possible to also have evidence on continuous variables. cplint on SWISH offers also the possibility of sampling arguments of goals, a kind of inference rarely considered but useful especially when the arguments are continuous variables. Finally, cplint on SWISH offers the possibility of graphing the results, for example by drawing the distribution of the sampled continuous arguments of goals.

Causal inference in cplint

Abstract cplint is a suite of programs for reasoning and learning with Probabilistic Logic Programming languages that follow the distribution semantics. In this paper we describe how we have extended cplint to perform causal reasoning. In particular, we consider Pearls do calculus for models where all the variables are measured. The two cplint modules for inference, PITA and MCINTYRE, have been extended for computing the effect of actions/interventions on these models. We also executed experiments comparing exact and approximate inference with conditional and causal queries, showing that causal inference is often cheaper than conditional inference.