arXiv subjects

Ana Ozaki

Publications and source records attributed to Ana Ozaki.

At least 19 recordsLinked to original sources

arXiv2026

BoxLitE: A Faithful Knowledge Base Embedding Based on Convex Optimization

Knowledge base (KB) embeddings aim at combining the capability of classical knowledge graph embeddings to generalize the information present in facts, the ABox, with conceptual knowledge represented in an ontology language, the TBox. Several authors have recently explored the idea of mapping concepts to convex regions in a vector space. This is useful to represent hierarchies, typically present in TBoxes, since more general concepts can be mapped to larger regions, containing those regions associated with more specific concepts. However, the power of convexity is rarely leveraged during the actual learning tasks. Here, we introduce BoxLitE, a KB embedding model for DL-Lite$^{\mathcal{H}}$ that allows for convex optimization. We show that for any satisfiable DL-Lite$^{\mathcal{H}}$ KB, there is a BoxLitE embedding that is a weakly faithful model. As a proof of concept, we show how to formulate the KB embedding task as a convex optimization problem and how to obtain embeddings with such desirable faithfulness properties.

cs.AI↗

arXiv2026

Semiring Provenance for Lightweight Description Logics

We investigate semiring provenance--a successful framework originally defined in the relational database setting--for description logics. In this context, the ontology axioms are annotated with elements of a commutative semiring and these annotations are propagated to the ontology consequences in a way that reflects how they are derived. We define a provenance semantics for a language that encompasses several lightweight description logics and show its relationships with semantics that have been defined for ontologies annotated with a specific kind of annotation (such as fuzzy degrees). We show that under some restrictions on the semiring, the semantics satisfies desirable properties (such as extending the semiring provenance defined for databases). We then focus on the well-known why-provenance, for which we study the complexity of problems related to the provenance of an assertion or a conjunctive query answer. Finally, we consider two more restricted cases which correspond to the so-called positive Boolean provenance and lineage in the database setting. For these cases, we exhibit relationships with well-known notions related to explanations in description logics and complete our complexity analysis. As a side contribution, we provide conditions on an $\mathcal{ELHI}_\bot$ ontology that guarantee tractable reasoning.

cs.LO↗

arXiv2026

Model Change for Description Logic Concepts

We consider the problem of modifying a description logic concept in light of models represented as pointed interpretations. We call this setting model change, and distinguish three main kinds of changes: eviction, which consists of only removing models; reception, which incorporates models; and revision, which combines removal with incorporation of models in a single operation. We introduce a formal notion of revision and argue that it does not reduce to a simple combination of eviction and reception, contrary to intuition. We provide positive and negative results on the compatibility of eviction and reception for EL and ALC description logic concepts and on the compatibility of revision for ALC concepts.

cs.LO↗

arXiv2025

Extracting PAC Decision Trees from Black Box Binary Classifiers: The Gender Bias Case Study on BERT-based Language Models

Decision trees are a popular machine learning method, known for their inherent explainability. In Explainable AI, decision trees can be used as surrogate models for complex black box AI models or as approximations of parts of such models. A key challenge of this approach is determining how accurately the extracted decision tree represents the original model and to what extent it can be trusted as an approximation of their behavior. In this work, we investigate the use of the Probably Approximately Correct (PAC) framework to provide a theoretical guarantee of fidelity for decision trees extracted from AI models. Based on theoretical results from the PAC framework, we adapt a decision tree algorithm to ensure a PAC guarantee under certain conditions. We focus on binary classification and conduct experiments where we extract decision trees from BERT-based language models with PAC guarantees. Our results indicate occupational gender bias in these models.

cs.AI↗

arXiv2025

On Middle Grounds for Preference Statements

In group decisions or deliberations, stakeholders are often confronted with conflicting opinions. We investigate a logic-based way of expressing such opinions and a formal general notion of a middle ground between stakeholders. Inspired by the literature on preferences with hierarchical and lexicographic models, we instantiate our general framework to the case where stakeholders express their opinions using preference statements of the form I prefer 'a' to 'b', where 'a' and 'b' are alternatives expressed over some attributes, e.g., in a trolley problem, one can express I prefer to save 1 adult and 1 child to 2 adults (and 0 children). We prove theoretical results on the existence and uniqueness of middle grounds. In particular, we show that, for preference statements, middle grounds may not exist and may not be unique. We also provide algorithms for deciding the existence and finding middle grounds.

cs.LO↗

arXiv2024

On the Power and Limitations of Examples for Description Logic Concepts

Labeled examples (i.e., positive and negative examples) are an attractive medium for communicating complex concepts. They are useful for deriving concept expressions (such as in concept learning, interactive concept specification, and concept refinement) as well as for illustrating concept expressions to a user or domain expert. We investigate the power of labeled examples for describing description-logic concepts. Specifically, we systematically study the existence and efficient computability of finite characterisations, i.e. finite sets of labeled examples that uniquely characterize a single concept, for a wide variety of description logics between EL and ALCQI, both without an ontology and in the presence of a DL-Lite ontology. Finite characterisations are relevant for debugging purposes, and their existence is a necessary condition for exact learnability with membership queries.

cs.LO↗

arXiv2024

Strong Faithfulness for ELH Ontology Embeddings

Ontology embedding methods are powerful approaches to represent and reason over structured knowledge in various domains. One advantage of ontology embeddings over knowledge graph embeddings is their ability to capture and impose an underlying schema to which the model must conform. Despite advances, most current approaches do not guarantee that the resulting embedding respects the axioms the ontology entails. In this work, we formally prove that normalized ${\cal ELH}$ has the strong faithfulness property on convex geometric models, which means that there is an embedding that precisely captures the original ontology. We present a region-based geometric model for embedding normalized ${\cal ELH}$ ontologies into a continuous vector space. To prove strong faithfulness, our construction takes advantage of the fact that normalized ${\cal ELH}$ has a finite canonical model. We first prove the statement assuming (possibly) non-convex regions, allowing us to keep the required dimensions low. Then, we impose convexity on the regions and show the property still holds. Finally, we consider reasoning tasks on geometric models and analyze the complexity in the class of convex geometric models used for proving strong faithfulness.

cs.LO↗

arXiv2024

Knowledge Base Embeddings: Semantics and Theoretical Properties

Research on knowledge graph embeddings has recently evolved into knowledge base embeddings, where the goal is not only to map facts into vector spaces but also constrain the models so that they take into account the relevant conceptual knowledge available. This paper examines recent methods that have been proposed to embed knowledge bases in description logic into vector spaces through the lens of their geometric-based semantics. We identify several relevant theoretical properties, which we draw from the literature and sometimes generalize or unify. We then investigate how concrete embedding methods fit in this theoretical framework.

cs.AI↗

arXiv2023

Rule Learning as Machine Translation using the Atomic Knowledge Bank

Machine learning models, and in particular language models, are being applied to various tasks that require reasoning. While such models are good at capturing patterns their ability to reason in a trustable and controlled manner is frequently questioned. On the other hand, logic-based rule systems allow for controlled inspection and already established verification methods. However it is well-known that creating such systems manually is time-consuming and prone to errors. We explore the capability of transformers to translate sentences expressing rules in natural language into logical rules. We see reasoners as the most reliable tools for performing logical reasoning and focus on translating language into the format expected by such tools. We perform experiments using the DKET dataset from the literature and create a dataset for language to logic translation based on the Atomic knowledge bank.

cs.CL↗

arXiv2023

Provenance for the Description Logic ELHr

We address the problem of handling provenance information in ELHr ontologies. We consider a setting recently introduced for ontology-based data access, based on semirings and extending classical data provenance, in which ontology axioms are annotated with provenance tokens. A consequence inherits the provenance of the axioms involved in deriving it, yielding a provenance polynomial as an annotation. We analyse the semantics for the ELHr case and show that the presence of conjunctions poses various difficulties for handling provenance, some of which are mitigated by assuming multiplicative idempotency of the semiring. Under this assumption, we study three problems: ontology completion with provenance, computing the set of relevant axioms for a consequence, and query answering.

cs.LO↗

arXiv2023

Learning Horn Envelopes via Queries from Large Language Models

We investigate an approach for extracting knowledge from trained neural networks based on Angluin's exact learning model with membership and equivalence queries to an oracle. In this approach, the oracle is a trained neural network. We consider Angluin's classical algorithm for learning Horn theories and study the necessary changes to make it applicable to learn from neural networks. In particular, we have to consider that trained neural networks may not behave as Horn oracles, meaning that their underlying target theory may not be Horn. We propose a new algorithm that aims at extracting the "tightest Horn approximation" of the target theory and that is guaranteed to terminate in exponential time (in the worst case) and in polynomial time if the target has polynomially many non-Horn examples. To showcase the applicability of the approach, we perform experiments on pre-trained language models and extract rules that expose occupation-based gender biases.

cs.LG↗

arXiv2023

Finite Based Contraction and Expansion via Models

We propose a new paradigm for Belief Change in which the new information is represented as sets of models, while the agent's body of knowledge is represented as a finite set of formulae, that is, a finite base. The focus on finiteness is crucial when we consider limited agents and reasoning algorithms. Moreover, having the input as arbitrary set of models is more general than the usual treatment of formulae as input. In this setting, we define new Belief Change operations akin to traditional expansion and contraction, and we identify the rationality postulates that emerge due to the finite representability requirement. We also analyse different logics concerning compatibility with our framework.

cs.LO↗

arXiv2023

Non-Normal Modal Description Logics (Extended Version)

Modal logics are widely used in multi-agent systems to reason about actions, abilities, norms, or epistemic states. Combined with description logic languages, they are also a powerful tool to formalise modal aspects of ontology-based reasoning over an object domain. However, the standard relational semantics for modalities is known to validate principles deemed problematic in agency, deontic, or epistemic applications. To overcome these difficulties, weaker systems of so-called non-normal modal logics, equipped with neighbourhood semantics that generalise the relational one, have been investigated both at the propositional and at the description logic level. We present here a family of non-normal modal description logics, obtained by extending ALC-based languages with non-normal modal operators. For formulas interpreted on neighbourhood models over varying domains, we provide a modular framework of terminating, correct, and complete tableau-based satisfiability checking algorithms in NExpTime. For a subset of these systems, we also consider a reduction to satisfiability on constant domain relational models. Moreover, we investigate the satisfiability problem in fragments obtained by disallowing the application of modal operators to description logic concepts, providing tight ExpTime complexity results.

cs.LO↗

arXiv2023

Verifying Properties of Tsetlin Machines

Tsetlin Machines (TsMs) are a promising and interpretable machine learning method which can be applied for various classification tasks. We present an exact encoding of TsMs into propositional logic and formally verify properties of TsMs using a SAT solver. In particular, we introduce in this work a notion of similarity of machine learning models and apply our notion to check for similarity of TsMs. We also consider notions of robustness and equivalence from the literature and adapt them for TsMs. Then, we show the correctness of our encoding and provide results for the properties: adversarial robustness, equivalence, and similarity of TsMs. In our experiments, we employ the MNIST and IMDB datasets for (respectively) image and sentiment classification. We discuss the results for verifying robustness obtained with TsMs with those in the literature obtained with Binarized Neural Networks on MNIST.

cs.LG↗

arXiv2023

Living Without Beth and Craig: Definitions and Interpolants in Description and Modal Logics with Nominals and Role Inclusions

The Craig interpolation property (CIP) states that an interpolant for an implication exists iff it is valid. The projective Beth definability property (PBDP) states that an explicit definition exists iff a formula stating implicit definability is valid. Thus, the CIP and PBDP reduce potentially hard existence problems to entailment in the underlying logic. Description (and modal) logics with nominals and/or role inclusions do not enjoy the CIP nor the PBDP, but interpolants and explicit definitions have many applications, in particular in concept learning, ontology engineering, and ontology-based data management. In this article we show that, even without Beth and Craig, the existence of interpolants and explicit definitions is decidable in description logics with nominals and/or role inclusions such as ALCO, ALCH and ALCHOI and corresponding hybrid modal logics. However, living without Beth and Craig makes this problem harder than entailment: the existence problems become 2ExpTime-complete in the presence of an ontology or the universal modality, and coNExpTime-complete otherwise. We also analyze explicit definition existence if all symbols (except the one that is defined) are admitted in the definition. In this case the complexity depends on whether one considers individual or concept names. Finally, we consider the problem of computing interpolants and explicit definitions if they exist and turn the complexity upper bound proof into an algorithm computing them, at least for description logics with role inclusions.

cs.LO↗

arXiv2022

Finding Common Ground for Incoherent Horn Expressions

Autonomous systems that operate in a shared environment with people need to be able to follow the rules of the society they occupy. While laws are unique for one society, different people and institutions may use different rules to guide their conduct. We study the problem of reaching a common ground among possibly incoherent rules of conduct. We formally define a notion of common ground and discuss the main properties of this notion. Then, we identify three sufficient conditions on the class of Horn expressions for which common grounds are guaranteed to exist. We provide a polynomial time algorithm that computes common grounds, under these conditions. We also show that if any of the three conditions is removed then common grounds for the resulting (larger) class may not exist.

cs.AI↗

arXiv2022

Reasoning in Non-normal Modal Description Logics

Non-normal modal logics, interpreted on neighbourhood models which generalise the usual relational semantics, have found application in several areas, such as epistemic, deontic, and coalitional reasoning. We present here preliminary results on reasoning in a family of modal description logics obtained by combining ALC with non-normal modal operators. First, we provide a framework of terminating, correct, and complete tableau algorithms to check satisfiability of formulas in such logics with the semantics based on varying domains. We then investigate the satisfiability problems in fragments of these languages obtained by restricting the application of modal operators to formulas only, and interpreted on models with constant domains, providing tight complexity results.

cs.LO↗

arXiv2022

On the Effectiveness of Knowledge Graph Embeddings: a Rule Mining Approach

We study the effectiveness of Knowledge Graph Embeddings (KGE) for knowledge graph (KG) completion with rule mining. More specifically, we mine rules from KGs before and after they have been completed by a KGE to compare possible differences in the rules extracted. We apply this method to classical KGEs approaches, in particular, TransE, DistMult and ComplEx. Our experiments indicate that there can be huge differences between the extracted rules, depending on the KGE approach for KG completion. In particular, after the TransE completion, several spurious rules were extracted.

cs.LG↗