arXiv subjects

Pierre Genevès

Publications and source records attributed to Pierre Genevès.

11 recordsLinked to original sources

arXiv2026

Beyond Accuracy: Measuring Logical Compliance of Predictive Models

Machine learning models are predominantly evaluated through predictive performance metrics such as ranking quality, prediction error, or classification accuracy. While these metrics effectively quantify how closely predictions match the ground truth, they do not assess whether model outputs respect predefined logical or domain-specific constraints. In high-stakes applications, including healthcare, finance, and autonomous systems, logical consistency can be as critical as predictive accuracy, yet no standard metric captures this dimension. We introduce the Rule Violation Score (RVS), a complementary evaluation metric that quantifies the extent to which a predictive model respects a given set of logical rules, independently of predictive accuracy. RVS treats hard rules (strict constraints) and soft rules (statistical regularities) differently, can be evaluated on any dataset and on any predictive model expressed over a relational vocabulary, and can be computed using SQL queries that are automatically generated for Horn rules. Beyond evaluating models, RVS can also evaluate the logical consistency of training datasets and help identify poorly defined rules. We evaluate RVS on three benchmarks covering knowledge graph link prediction and relational regression, including rule-based, embedding-based, and neuro-symbolic predictive models. Our results demonstrate that two models achieving comparable predictive accuracy can exhibit substantially different levels of logical compliance, revealing differences in model behavior that standard metrics fail to capture.

cs.AI↗

arXiv2026

Towards Multiparty Session Types for Highly-Concurrent and Fault-Tolerant Web Applications

Modern web applications combine persistent state updates, concurrent interactions, and unreliable communication with external services. Failures such as timeouts can occur after partial state changes, producing temporary inconsistencies whose resolution depends on liveness properties that are often not verified in practice. Although formal methods offer rigorous guarantees for reasoning about complex software, they remain rarely adopted in enterprise settings due to their perceived complexity and lack of practical automation. Multiparty Session Types (MPST) offer strong guarantees for communication safety, yet they do not account for the interplay between state evolution, dynamic workflow structure, and failure behaviour that are essential for reasoning about the correctness of real web applications. This paper introduces a global-type framework that equips MPST with explicit failure semantics and dynamic participation. We define the syntax and operational semantics of these enriched global types and establish core properties, including coherence preservation. This foundation enables formal reasoning about communications in web applications where failures may occur, and lays the groundwork for future stateful extensions and automated verification of liveness properties.

cs.PL↗

arXiv2026

Optimizing Relational Queries over Array-Valued Data in Columnar Systems

Modern analytical workloads increasingly combine relational data with array-valued attributes. While columnar database systems efficiently process such workloads, their ability to optimize queries that interleave relational operators with array manipulations remains limited. This paper introduces A3D-RA, an extended relational algebra supporting array-valued attributes, together with a comprehensive framework for algebraic reasoning and optimization. We formalize its data model and semantics, develop a complete set of equivalence-preserving transformation rules capturing pairwise interactions between relational and array operators, and propose a plan enumeration strategy with an optimality guarantee that remains polynomial in all non-join operators. We design A3D-RA as a modular, backend-independent optimization layer that can be instantiated over existing analytical database systems. Experimental results across three high-performance engines on a real-world workload show consistent performance gains enabled by the proposed algebraic optimization layer.

cs.DB↗

arXiv2025

Distributed Evaluation of Graph Queries using Recursive Relational Algebra

We present a system called Dist-$μ$-RA for the distributed evaluation of recursive graph queries. Dist-$μ$-RA builds on the recursive relational algebra and extends it with evaluation plans suited for the distributed setting. The goal is to offer expressivity for high-level queries while providing efficiency at scale and reducing communication costs. Experimental results on both real and synthetic graphs show the effectiveness of the proposed approach compared to existing systems.

cs.DB↗

arXiv2025

Schema-Based Query Optimisation for Graph Databases

Recursive graph queries are increasingly popular for extracting information from interconnected data found in various domains such as social networks, life sciences, and business analytics. Graph data often come with schema information that describe how nodes and edges are organized. We propose a type inference mechanism that enriches recursive graph queries with relevant structural information contained in a graph schema. We show that this schema information can be useful in order to improve the performance when evaluating acylic recursive graph queries. Furthermore, we prove that the proposed method is sound and complete, ensuring that the semantics of the query is preserved during the schema-enrichment process.

cs.DB↗

arXiv2025

On Scaling Neurosymbolic Programming through Guided Logical Inference

Probabilistic neurosymbolic learning seeks to integrate neural networks with symbolic programming. Many state-of-the-art systems rely on a reduction to the Probabilistic Weighted Model Counting Problem (PWMC), which requires computing a Boolean formula called the logical provenance.However, PWMC is \\#P-hard, and the number of clauses in the logical provenance formula can grow exponentially, creating a major bottleneck that significantly limits the applicability of PNL solutions in practice.We propose a new approach centered around an exact algorithm DPNL, that enables bypassing the computation of the logical provenance.The DPNL approach relies on the principles of an oracle and a recursive DPLL-like decomposition in order to guide and speed up logical inference.Furthermore, we show that this approach can be adapted for approximate reasoning with $ε$ or $(ε, δ)$ guarantees, called ApproxDPNL.Experiments show significant performance gains.DPNL enables scaling exact inference further, resulting in more accurate models.Further, ApproxDPNL shows potential for advancing the scalability of neurosymbolic programming by incorporating approximations even further, while simultaneously ensuring guarantees for the reasoning process.

cs.AI↗

arXiv2024

Efficient Enumeration of Recursive Plans in Transformation-based Query Optimizers

Query optimizers built on the transformation-based Volcano/Cascades framework are used in many database systems. Transformations proposed earlier on the logical query dag (LQDAG) data structure, which is key in such a framework, focus only on recursion-free queries. In this paper, we propose the recursive logical query dag (RLQDAG) which extends the LQDAG with the ability to capture and transform recursive queries, leveraging recent developments in recursive relational algebra. Specifically, this extension includes: (i) the ability of capturing and transforming sets of recursive relational terms thanks to (ii) annotated equivalence nodes used for guiding transformations that are more complex in the presence of recursion; and (iii) RLQDAG rewrite rules that transform sets of subterms in a grouped manner, instead of transforming individual terms in a sequential manner; and that (iv) incrementally update the necessary annotations. Core concepts of the RLQDAG are formalized using a syntax and formal semantics with a particular focus on subterm sharing and recursion. The result is a clean generalization of the LQDAG transformation-based approach, enabling more efficient explorations of plan spaces for recursive queries. An implementation of the proposed approach shows significant performance gains compared to the state-of-the-art.

cs.DB↗

arXiv2023

Knowledge Enhanced Graph Neural Networks for Graph Completion

Graph data is omnipresent and has a wide variety of applications, such as in natural science, social networks, or the semantic web. However, while being rich in information, graphs are often noisy and incomplete. As a result, graph completion tasks, such as node classification or link prediction, have gained attention. On one hand, neural methods, such as graph neural networks, have proven to be robust tools for learning rich representations of noisy graphs. On the other hand, symbolic methods enable exact reasoning on graphs.We propose Knowledge Enhanced Graph Neural Networks (KeGNN), a neuro-symbolic framework for graph completion that combines both paradigms as it allows for the integration of prior knowledge into a graph neural network model.Essentially, KeGNN consists of a graph neural network as a base upon which knowledge enhancement layers are stacked with the goal of refining predictions with respect to prior knowledge.We instantiate KeGNN in conjunction with two state-of-the-art graph neural networks, Graph Convolutional Networks and Graph Attention Networks, and evaluate KeGNN on multiple benchmark datasets for node classification.

cs.AI↗

arXiv2021

Constrained Differentially Private Federated Learning for Low-bandwidth Devices

Federated learning becomes a prominent approach when different entities want to learn collaboratively a common model without sharing their training data. However, Federated learning has two main drawbacks. First, it is quite bandwidth inefficient as it involves a lot of message exchanges between the aggregating server and the participating entities. This bandwidth and corresponding processing costs could be prohibitive if the participating entities are, for example, mobile devices. Furthermore, although federated learning improves privacy by not sharing data, recent attacks have shown that it still leaks information about the training data. This paper presents a novel privacy-preserving federated learning scheme. The proposed scheme provides theoretical privacy guarantees, as it is based on Differential Privacy. Furthermore, it optimizes the model accuracy by constraining the model learning phase on few selected weights. Finally, as shown experimentally, it reduces the upstream and downstream bandwidth by up to 99.9% compared to standard federated learning, making it practical for mobile systems.

cs.CR↗

arXiv2020

Compression Boosts Differentially Private Federated Learning

Federated Learning allows distributed entities to train a common model collaboratively without sharing their own data. Although it prevents data collection and aggregation by exchanging only parameter updates, it remains vulnerable to various inference and reconstruction attacks where a malicious entity can learn private information about the participants' training data from the captured gradients. Differential Privacy is used to obtain theoretically sound privacy guarantees against such inference attacks by noising the exchanged update vectors. However, the added noise is proportional to the model size which can be very large with modern neural networks. This can result in poor model quality. In this paper, compressive sensing is used to reduce the model size and hence increase model quality without sacrificing privacy. We show experimentally, using 2 datasets, that our privacy-preserving proposal can reduce the communication costs by up to 95% with only a negligible performance penalty compared to traditional non-private federated learning schemes.

cs.LG↗

arXiv2008

Ensuring Query Compatibility with Evolving XML Schemas

During the life cycle of an XML application, both schemas and queries may change from one version to another. Schema evolutions may affect query results and potentially the validity of produced data. Nowadays, a challenge is to assess and accommodate the impact of theses changes in rapidly evolving XML applications. This article proposes a logical framework and tool for verifying forward/backward compatibility issues involving schemas and queries. First, it allows analyzing relations between schemas. Second, it allows XML designers to identify queries that must be reformulated in order to produce the expected results across successive schema versions. Third, it allows examining more precisely the impact of schema changes over queries, therefore facilitating their reformulation.

cs.PL↗