Probabilistic soft logic

Probabilistic Soft Logic (PSL) is a statistical relational learning (SRL) framework for modeling probabilistic and relational domains.

Uses your system voice; voice and rate in settings.

Snapshot

≈ 6 min read

Names in other languages

No names in other languages are held for this entry yet.

Article bodies are English and are not machine-translated. A name in another language is not a translation of the article.

Probabilistic Soft Logic (PSL) is a statistical relational learning (SRL) framework for modeling probabilistic and relational domains. It is applicable to a variety of machine learning problems, such as collective classification, entity resolution, link prediction, and ontology alignment.

PSL combines two tools: first-order logic, with its ability to succinctly represent complex phenomena, and probabilistic graphical models, which capture the uncertainty and incompleteness inherent in real-world knowledge. More specifically, PSL uses "soft" logic as its logical component and Markov random fields as its statistical model.

PSL provides sophisticated inference techniques for finding the most likely answer (i.e. the maximum a posteriori (MAP) state). The "softening" of the logical formulas makes inference a polynomial time operation rather than an NP-hard operation.

Similar(A, B) & HasLabel(A, X) -> HasLabel(B, X) This rule can be interpreted to mean: If A and B are similar and A has the label X, then there is evidence that B also has the label X. Arithmetic rules are relations of two linear combinations of atoms. Restricting each side to a linear combination ensures that the resulting potential is convex.

The following relational operators are supported: = , , and >= . Similar(A, B) = Similar(B, A) This rule encodes the notion that similarity is symmetric in this model. A commonly used feature of arithmetic rules is the summation operation.

The summation operation can be used to aggregate multiple atoms. When used, the atom is replaced with the sum of all possible atoms where the non-summation variables are fixed. Summation variables are made by prefixing a variable with a + .

Fox example: HasLabel(A, +X) = 1.0 If the possible values for X are label1, label2, and label3, then the above rule is equivalent to: HasLabel(A, 'label1') + HasLabel(A, 'label2') + HasLabel(A, 'label3') = 1.0 Both of these rules force the sum of all possible labels for an entity to sum to 1.0. This type of rule is especially useful for collective classification problems, where only one class can be selected.

An HL-MRF determines a density function over a set of continuous variables \mathbf{y} = (y_1, \cdots , y_n) with joint domain [0, 1]^n using set of evidence \mathbf{x} = (x_1, \cdots , x_m) , weights \mathbf{w} = (w_1, \cdots , w_k) , and potential functions \mathbf{\phi} = (\phi_1, \cdots , \phi_k) of the form \mathbf{\phi_i (\mathbf{x}, \mathbf{y})} = \max(\ell_i (\mathbf{x}, \mathbf{y}), 0)^{d_i} where \ell_i is a linear function and d_i \in \{1,2\} . The conditional distribution of \mathbf{y} given the observed data \mathbf{x} is defined as P(\mathbf{y} | \mathbf{x}) = \frac{1}{Z(\mathbf{y})} \exp(\sum_{i = 1}^{k} w_i \phi_i (\mathbf{x}, \mathbf{y})) Where Z(\mathbf{y}) = \int_{\mathbf{y}} \exp(\sum_{i = 1}^{k} w_i \phi_i (\mathbf{x}, \mathbf{y})) is the partition function.

This density is a logarithmically convex function, and thus the common inference task in PSL of finding a maximum a posteriori estimation of the joint state of \mathbf{y} is a convex problem. This allows inference in PSL to be achievable in polynomial-time.

For example, if we had the following constants in the data for representing people: \{Alice, Bob\} and the following constant for movies: \{Avatar\} , and we provided PSL with the predicate data \{rating(Alice, Avatar) = 0.8\} and rating(\cdot) was labeled closed, then PSL would assume that \{rating(Bob, Avatar) = 0\} even though this data was never explicitly provided to the system. If a predicate is labeled as open, then PSL does not make the closed-world assumption.

Instead, PSL will attempt to collectively infer the unobserved instances. Grounding predicates in PSL is the process of making all possible substitutions of the variables in each predicate with the existing constants in the data, resulting in a collection of ground atoms, \mathbf{y} = \{y_1, \cdots, y_n\} .

Then, all possible substitutions of the ground atoms for the predicates in the rules are made to create ground rules. Each of the ground rules are interpreted as either potentials or hard constraints in the induced HL-MRF. A logical rule is translated as a continuous relaxation of Boolean connectives using Łukasiewicz logic.

A ground logical rule is transformed into its disjunctive normal form. Let I^{+} be the set of indices of the variables that correspond to atoms that are not negated, and, likewise I^{-} the set of indices corresponding to atoms that are negated, in the disjunctive clause.

Then the logical rule maps to the inequality: 1 - \sum_{i \in I^+} y_i - \sum_{i \in I^-}(1 - y_i) \leq 0 If the logical rule is weighted with a weight w and exponentiated with d \in \{1, 2\} , then the potential \phi(\mathbf{y}) = \Big ( \max \Big \{ 1 - \sum_{i \in I^+} y_i - \sum_{i \in I^-}(1 - y_i), 0\Big \} \Big )^{d} is added to the HL-MRF with a weight parameter of w . An arithmetic rule is manipulated to \ell(\mathbf{y}) \leq 0 and the resulting potential takes the form \phi(\mathbf{y}) = (\max \{\ell(\mathbf{y}), 0\})^{d} . /* People living in the same location are more likely to know one another. */ 20: Lived(P1, L) & Lived(P2, L) & (P1 != P2) -> Knows(P1, P2) ^2 /* People who have not lived in the same location are not likely to know one another. */ 5: Lived(P1, L1) & Lived(P2, L2) & (P1 != P2) & (L1 != L2) -> !Knows(P1, P2) ^2 /* Two people with similar interests are more likely to know one another. */ 10: Likes(P1, X) & Likes(P2, X) & (P1 != P2) -> Knows(P1, P2) ^2 /* People in the same circles tend to know one another (transitivity). */ 5: Knows(P1, P2) & Knows(P2, P3) & (P1 != P3) -> Knows(P1, P3) ^2 /* Knowing one another is symmetric. */ Knows(P1, P2) = Knows(P2, P1) . /* By default, assume that two arbitrary people do not know one another (negative prior). */ 5: !Knows(P1, P2) ^2

Text derived from https://en.wikipedia.org/w/index.php?oldid=1339379166 under CC BY-SA 4.0; modifications have been made.

Provenance

snapshot enwiki-001339379166 · captured 2026-02-20 · not yet sealed in the ledger · verify the chain · version history

Cite this page as: encyclopegia.org, “Probabilistic soft logic”, snapshot enwiki-001339379166. https://encyclopegia.org/a/probabilistic-soft-logic

Sources cited in this revision

13 named source(s) extracted from the cited revision. A source with no author is shown by its publication; nothing is invented.