Dedicated Tabling for a Probabilistic Setting

@inproceedings{Mantadelis2010DedicatedTF,
  title={Dedicated Tabling for a Probabilistic Setting},
  author={Theofrastos Mantadelis and Gerda Janssens},
  booktitle={ICLP},
  year={2010}
}
ProbLog is a probabilistic framework that extends Prolog with probabilistic facts. To compute the probability of a query, the complete SLD proof tree of the query is collected as a sum of products. ProbLog applies advanced techniques to make this feasible and to assess the correct probability. Tabling is a well-known technique to avoid repeated subcomputations and to terminate loops. We investigate how tabling can be used in ProbLog. The challenge is that we have to reconcile tabling with the… 

Figures and Tables from this paper

Nesting Probabilistic Inference
TLDR
In order to support nested probabilistic inference in ProbLog, the notion of a parametrised ProbLog engine is proposed and several extensions of ProbLog such as meta-calls, negation, and answers of probabilism goals are realised.
Using Iterative Deepening for Probabilistic Logic Inference
TLDR
An iterative deepening algorithm is presented which handles cycles and produces solutions to problems that previously ProbLog was not able to solve and is able to get the exact result for the same or one lower scaling factor.
The Most Probable Explanation for Probabilistic Logic Programs with Annotated Disjunctions
TLDR
This work proposes a new encoding of annotated disjunctions which allows correct MARG and MPE and explores from both theoretical and experimental perspective the trade-off between the encoding suitable only for MARG inference and the newly proposed general approach.
Tabling for infinite probability computation
TLDR
This paper uses PRISM, a logic-based probabilistic modeling language with a tabling mechanism, to generalize prefix probability computation for PCFGs to Probabilistic logic programs, and opens a way to logic- based probabilism modeling of cyclic dependencies.
Inference in Probabilistic Logic Programs using Weighted CNF's
TLDR
This paper develops efficient inference algorithms for classical probabilistic inference tasks such as MAP and computing marginals based on a conversion of the Probabilistic logic program and the query and evidence to a weighted CNF formula.
Infinite probability computation by cyclic explanation graphs
TLDR
The general approach to prefix probability computation through tabling not only allows to deal with non-probabilistic context-free grammars such as probabilistic left-corner Grammars but has applications such as plan recognition and Probabilistic model checking and makes it possible to compute probability for probabilism models describing cyclic relations.
Compacting Boolean Formulae for Inference in Probabilistic Logic Programming
TLDR
This work identifies 7 Boolean subformulae patterns that can be used to re-write Boolean formulae and implements an algorithm with polynomial complexity which detects and compacts 6 of these patterns, improving knowledge compilation and consecutively the overall inference performance.
Well–definedness and efficient inference for probabilistic logic programming under the distribution semantics
TLDR
This paper presents the algorithm “Probabilistic Inference with Tabling and Answer subsumption” (PITA) that computes the probability of queries by transforming a probabilistic program into a normal program and then applying SLG resolution with answer subsumption.
ProbLog2: From probabilistic programming to statistical relational learning
TLDR
This work argues that the main mechanism behind ProbLog2 is a conversion of the given program to a weighted Boolean formula that can be applied with certain restrictions to other probabilistic programming languages such as Church and Figaro.
aspmc: An Algebraic Answer Set Counter
We report about aspmc, which is the working prototype implementation of recent advances in Algebraic Answer Set Counting (AASC). AASC refers to counting the answer sets of a normal answer set program
...
1
2
3
...

References

SHOWING 1-10 OF 18 REFERENCES
Tabling relevant parts of SLD proofs for ground goals in a probabilistic setting
TLDR
This work investigates how the idea of tabling can be used in ProbLog: the challenge is that only the relevant probabilistic parts of the SLD proof tree are required, and outlines a direction to tabling ProbLog programs.
Trading memory for answers: Towards tabling ProbLog
TLDR
Extensions of exact inference to deal with negated ground subgoals in clause bodies are discussed that broaden the class of problems that can be handled and can speed up inference in domains where dierent proofs of a query share manySubgoals.
ProbLog: A Probabilistic Prolog and its Application in Link Discovery
TLDR
The key contribution of this paper is the introduction of an effective solver for computing success probabilities, which essentially combines SLD-resolution with methods for computing the probability of Boolean formulae.
On the Efficient Execution of ProbLog Programs
TLDR
Algorithms that allow the efficient execution of queries are introduced, their implementation on top of the YAP-Prolog system is discussed, and their performance in the context of large networks of biological entities is evaluated.
Speculative Beats Conservative Justification
TLDR
This paper introduces the concept of a speculative justifier, a justifier that select literals that can possibly provide evidence for the goal's truth, assume that their truth values correspond to thegoal's and proceed to build a justification for each of them.
Tabled Evaluation with Delaying for General Logic Programs
TLDR
The semantics of predicate logic as a programming language, Magic sets and bottom-up evaluation of well-founded models, and top-down query evaluation procedure for stratiied databases are reviewed.
XSB: A System for Effciently Computing WFS
TLDR
The well-founded model provides a natural and robust semantics for logic programs with negative literals in rule bodies in the SLG-WAM of XSB, a full Prolog system that closely approximates the ISO standard.
Online Justification for Tabled Logic Programs
TLDR
This paper introduces online justification, based on program transformation, to efficiently construct the evidence during query evaluation, while adding little overhead to the evaluation itself, to demonstrate the scalability of online justification.
A Statistical Learning Method for Logic Programs with Distribution Semantics
TLDR
The distribution semantics is a straightforward generalization of the traditional least model semantics and can capture semantics of diverse information processing systems ranging from Bayesian networks to Hidden Markov models to Boltzmann machines in a single framework with mathematical rigor.
Parameter Learning of Logic Programs for Symbolic-Statistical Modeling
TLDR
A logical/mathematical framework for statistical parameter learning of parameterized logic programs, i.e. definite clause programs containing probabilistic facts with a parameterized distribution, and a new EM algorithm that can significantly outperform the Inside-Outside algorithm.
...
1
2
...