Probabilistic Soft Logic for Semantic Textual Similarity

Abstract

Probabilistic Soft Logic (PSL) is a recently developed framework for probabilistic logic. We use PSL to combine logical and distributional representations of natural-language meaning, where distributional information is represented in the form of weighted inference rules. We apply this framework to the task of Semantic Textual Similarity (STS) (i.e. judging the semantic similarity of naturallanguage sentences), and show that PSL gives improved results compared to a previous approach based on Markov Logic Networks (MLNs) and a purely distributional approach.

UTexas: Natural Language Semantics using Distributional Semantics and Probabilistic Logic

“…We showed in Beltagy et al (2014a) how to perform the STS task using PSL. PSL does not work "out of the box" for STS, because Lukasiewicz's equation for the conjunction is very restrictive.…”

“…Our system builds on our prior work (Beltagy et al, 2013;Beltagy et al, 2014a;Beltagy and Mooney, 2014;Beltagy et al, 2014b). We use Boxer (Bos, 2008), a wide-coverage semantic analysis tool to map natural sentences to logical form.…”

Semantic Parsing using Distributional Semantics and Probabilistic Logic

“…PSL does not work "out of the box" for STS, because Lukasiewicz's equation for the conjunction is very restrictive. We addressed this problem (Beltagy et al, 2014) For each STS pair of sentences S 1 , S 2 , we run PSL twice, once where E = S 1 , Q = S 2 and another where E = S 2 , Q = S 1 , and output the two scores. The final similarity score is produced from a regressor trained to map the two PSL scores to the overall similarity score.…”

“…They are Statistical Relational Learning (SRL) techniques (Getoor and Taskar, 2007) that combine logical and statistical knowledge in one uniform framework, and provide a mechanism for coherent probabilistic inference. We implemented this semantic parser (Beltagy et al, 2013;Beltagy et al, 2014) and used it to perform two tasks that require deep semantic analysis, Recognizing Textual Entailment (RTE), and Semantic Textual Similarity (STS).…”

Statistical Relational Learning with Soft Quantifiers

“…Intuitively, value 0 means false and value 1 means true, while any value v ∈ [0, 1] represents a partial degree of truth. PSL has been used in various domains with promising results, including trust propagation [12], drug-target interaction prediction [8], knowledge graph identification [23], semantic textual similarity computation [4] and sentiment analysis in a social network [27], among many others. Reasoning with continuous values has also been addressed in fuzzyDL [5], however reasoning is not as efficient as in PSL.…”