Probabilistic soft logic: Difference between revisions

Revision as of 15:14, 8 December 2017

Revision as of 07:04, 27 March 2018

Line 1:

'''Probabilistic soft logic (PSL)''' is a [[Statistical relational learning|SRL]] framework for collective, [[probabilistic reasoning]] in relational domains. PSL uses [[first order logic]] rules as a template language for [[graphical model]]s over [[random variable]]s with soft truth values from the interval [0,1].

PSL uses [[first order logic]] rules as a template language for [[graphical model]]s over [[random variable]]s with soft truth values from the interval&nbsp;[0,1]<ref>{{cite journal |last1=Bach |first1=Stephen |last2=Broecheler |first2=Matthias |last3=Huang |first3=Bert |last4=Getoor |first4=Lise |date=2017 |title=Hinge-Loss Markov Random Fields and Probabilistic Soft Logic |journal=Journal of Machine Learning Research (JMLR) |volume=18 |pages=1-67}}</ref>.

== Description ==

In recent years there has been a rise in the approaches that combine [[graphical model]]s and [[first-order logic]] to allow the development of complex probabilistic models with relational structures.
A notable example of such approaches is [[Markov logic network]]s (MLNs).<ref>{{cite book|last1=Getoor|first1=Lise|last2=Taskar|first2=Ben|title=Introduction to Statistical Relational Learning|date=12 Oct 2007|publisher=MIT Press|isbn=0262072882}}</ref>
Like MLNs PSL is a modelling language (with an accompanying implementation<ref>{{cite web|url=https://github.com/linqs/psl|title=GitHub repository|accessdate=1626 OctoberMarch 20142018}}</ref>) for learning and predicting in relational domains.
Unlike MLNs, PSL uses soft truth values for predicates in an interval between [0,1].
This allows for the integrationunderlying ofinference similarityto functionsbe insolved the into models. This is useful in problems suchquickly as [[ontologya mapping]]{{dn|date=December 2017}} and [[entity resolution]]. Also, in PSL the formula syntax is restricted to rules withconvex conjunctiveoptimization bodiesproblem.

This is useful in problems such as [[collective classification]], [[link prediction]], [[social network]] modelling, and [[record linkage|object identification/entity resolution/record linkage]].

== See also ==

Line 10 ⟶ 16:

* [[Fuzzy logic]]

== References ==

{{Reflist}}

== External links ==

* [httphttps://psl.umiacslinqs.umd.edu/org Probabilistic soft logic main web page]

* [https://github.com/linqs/psl Official PSL implementation in [[Groovy (programming language)|Groovy]]]

* [httphttps://psllinqs.umiacssoe.umducsc.edu/publications.phpbiblio A list of publications about PSL]

* [https://www.youtube.com/channel/UCJjzqRLiAIa3qENUkzK0zMA Video lectures about PSL inon Youtube]

[[Category:Bayesian statistics]]