Author name cluster

João Leite 0001

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

14 papers

1 author row

NeSy Conference 2025 Conference Paper

Concept Probing: Where to Find Human-Defined Concepts

Manuel de Sousa Ribeiro
Afonso Leote
João Leite 0001

Concept probing has recently gained popularity as a way for humans to peek into what is encoded within artificial neural networks. In concept probing, additional classifiers are trained to map the internal representations of a model into human-defined concepts of interest. However, the performance of these probes is highly dependent on the internal representations they probe from, making identifying the appropriate layer to probe an essential task. In this paper, we propose a method to automatically identify which layer’s representations in a neural network model should be considered when probing for a given human-defined concept of interest, based on how informative and regular the representations are with respect to the concept. We validate our findings through an exhaustive empirical analysis over different neural network models and datasets.

Details

ECAI Conference 2025 Conference Paper

On the Performance of Concept Probing: The Influence of the Data

Manuel de Sousa Ribeiro
Afonso Leote
João Leite 0001

Concept probing has recently garnered increasing interest as a way to help interpret artificial neural networks, dealing both with their typically large size and their subsymbolic nature, which ultimately renders them unfeasible for direct human interpretation. Concept probing works by training additional classifiers to map the internal representations of a model into human-defined concepts of interest, thus allowing humans to peek inside artificial neural networks. Research on concept probing has mainly focused on the model being probed or the probing model itself, paying limited attention to the data required to train such probing models. In this paper, we address this gap. Focusing on concept probing in the context of image classification tasks, we investigate the effect of the data used to train probing models on their performance. We also make available concept labels for two widely used datasets.

Details

NMR Workshop 2024 Conference Paper

Abstract Dialectical Frameworks are Boolean Networks

Jesse Heyninck
Matthias Knorr 0001
João Leite 0001

Abstract dialectical frameworks are a unifying model of formal argumentation, where argumentative relations between arguments are represented by assigning acceptance conditions to atomic arguments. Their generality allows them to cover a number of different approaches with varying forms of representing the argumentation structure. Boolean regulatory networks are used to model the dynamics of complex biological processes, taking into account the interactions of biological compounds, such as proteins or genes. These models have proven highly useful for comprehending such biological processes, allowing to reproduce known behaviour and testing new hypotheses and predictions in silico, for example in the context of new medical treatments. While both these approaches stem from entirely different communities, it turns out that there are striking similarities in their appearence. In this paper, we study the relation between these two formalisms revealing their communalities as well as their differences, and introducing a correspondence that allows to establish novel results for the individual formalisms.

Details

JELIA Conference 2021 Conference Paper

On Syntactic Forgetting Under Uniform Equivalence

Ricardo Gonçalves 0001
Tomi Janhunen
Matthias Knorr 0001
João Leite 0001

Abstract Forgetting in Answer Set Programming (ASP) aims at reducing the language of a logic program without affecting the consequences over the remaining language. It has recently gained interest in the context of modular ASP where it allows simplifying a program of a module, making it more declarative, by omitting auxiliary atoms or hiding certain atoms/parts of the program not to be disclosed. Unlike for arbitrary programs, it has been shown that forgetting for modular ASP can always be applied, for input, output and hidden atoms, and preserve all dependencies over the remaining language (in line with uniform equivalence). However, the definition of the result is based solely on a semantic characterization in terms of HT-models. Thus, computing an actual result is a complicated process and the result commonly bears no resemblance to the original program, i. e. , we are lacking a corresponding syntactic operator. In this paper, we show that there is no forgetting operator that preserves uniform equivalence (modulo the forgotten atoms) between the given program and its forgetting result by only manipulating the rules of the original program that contain the atoms to be forgotten. We then present a forgetting operator that preserves uniform equivalence and is syntactic whenever this is suitable. We also introduce a special class of programs, where syntactic forgetting is always possible, and as a complementary result, establish it as the largest known class where forgetting while preserving all dependencies is always possible.

Details

JELIA Conference 2021 Conference Paper

Tractable Reasoning Using Logic Programs with Intensional Concepts

Jesse Heyninck
Ricardo Gonçalves 0001
Matthias Knorr 0001
João Leite 0001

Abstract Recent developments triggered by initiatives such as the Semantic Web, Linked Open Data, the Web of Things, and geographic information systems resulted in the wide and increasing availability of machine-processable data and knowledge in the form of data streams and knowledge bases. Applications building on such knowledge require reasoning with modal and intensional concepts, such as time, space, and obligations, that are defeasible. E. g. , in the presence of data streams, conclusions may have to be revised due to newly arriving information. The current literature features a variety of domain-specific formalisms that allow for defeasible reasoning using specific intensional concepts. However, many of these formalisms are computationally intractable and limited to one of the mentioned application domains. In this paper, we define a general method for obtaining defeasible inferences over intensional concepts, and we study conditions under which these inferences are computable in polynomial time.

Details

JELIA Conference 2016 Conference Paper

Forgetting in ASP: The Forgotten Properties

Ricardo Gonçalves 0001
Matthias Knorr 0001
João Leite 0001

Abstract Many approaches for forgetting in Answer Set Programming (ASP) have been proposed in recent years, in the form of specific operators, or classes of operators, following different principles and obeying different properties. A recently published comprehensive overview of existing operators and properties provided a uniform picture of the landscape, including many novel (even surprising) results on relations between properties and operators. Yet, this overview largely missed an additional set properties for forgetting, proposed by Wong, and this paper aims to close this gap. It turns out that, while some of these properties are closely related to the properties previously studied, four of them are distinct providing novel results and insights, further strengthening established relations between existing operators.

Details

JELIA Conference 2016 Conference Paper

Inconsistency Management in Reactive Multi-context Systems

Gerhard Brewka
Stefan Ellmauthaler
Ricardo Gonçalves 0001
Matthias Knorr 0001
João Leite 0001
Jörg Pührer

Abstract We address the problem of global inconsistency in reactive multi-context systems (rMCSs), a framework for reactive reasoning in the presence of heterogeneous knowledge sources that can deal with continuous input streams. Their semantics is given in terms of equilibria streams. The occurrence of inconsistencies, where rMCSs fail to have an equilibria stream, can render the entire system useless. We discuss various methods for handling this problem, following different strategies such as repairing the rMCS, or even relaxing the notion of equilibria stream so that it can go through inconsistent states.

Details

ECAI Conference 2016 Conference Paper

You Can't Always Forget What You Want: On the Limits of Forgetting in Answer Set Programming

Ricardo Gonçalves 0001
Matthias Knorr 0001
João Leite 0001

Selectively forgetting information while preserving what matters the most is becoming an increasingly important issue in many areas, including in knowledge representation and reasoning. Depending on the application at hand, forgetting operators are defined to obey different sets of desirable properties. Of the myriad of desirable properties discussed in the context of forgetting in Answer Set Programming, strong persistence, which imposes certain conditions on the correspondence between the answer sets of the program pre-and post-forgetting, and a certain independence from non-forgotten atoms, seems to best capture its essence, and be desirable in general. However, it has remained an open problem whether it is always possible to forget a set of atoms from a program while obeying strong persistence. In this paper, after showing that it is not always possible to forget a set of atoms from a program while obeying this property, we move forward and precisely characterise what can and cannot be forgotten from a program, by presenting a necessary and sufficient criterion. This characterisation allows us to draw some important conclusions regarding the existence of forgetting operators for specific classes of logic programs, to characterise the class of forgetting operators that achieve the correct result whenever forgetting is possible, and investigate the related question of determining what we can forget from some specific logic program.

Details

ECAI Conference 2014 Conference Paper

Evolving Multi-Context Systems

Ricardo Gonçalves 0001
Matthias Knorr 0001
João Leite 0001

Managed Multi-Context Systems (mMCSs) provide a general framework for integrating knowledge represented in heterogeneous KR formalisms. However, mMCSs are essentially static as they were not designed to run in a dynamic scenario. In this paper, we introduce evolving Multi-Context Systems (eMCSs), a general and flexible framework which inherits from mMCSs the ability to integrate knowledge represented in heterogeneous KR formalisms, and at the same time is able to both react to, and reason in the presence of commonly temporary dynamic observations, and evolve by incorporating new knowledge. We show that eMCSs are indeed very general and expressive enough to capture several existing KR approaches that model dynamics of knowledge.

Details

ECAI Conference 2014 Conference Paper

On the Efficient Implementation of Social Abstract Argumentation

Marco Correia
Jorge Cruz 0001
João Leite 0001

In this paper we present a novel iterative algorithm - the Iterative Successive Substitution (ISS) - to efficiently approximate the models of debates structured according to Social Abstract Argumentation [10]. Classical iterative algorithms such as the Iterative Newton-Raphson (INR) and the Iterative Fixed-point (IFP) don't always converge and, when they do, usually take too long to be effective. We analytically prove convergence of ISS, and empirically show that, even when INR and IFP converge, ISS always outperforms them, often by several orders of magnitude. The ISS is able to approximate the models of complex debates with thousands of arguments in well under a second, often in under one tenth of a second, making it comfortably suitable for its purpose. Additionally, we present a small modification to ISS that, with a negligible overhead, takes advantage of the topological structure of certain debates to significantly increase convergence times.

Details

JELIA Conference 2012 Conference Paper

A Unifying Perspective on Knowledge Updates

Martin Slota
João Leite 0001

Abstract We introduce an abstract update framework based on viewing a knowledge base as the set of sets of models of its elements and performing updates by introducing additional interpretations – exceptions – to the sets of models of elements of the original knowledge base. In [36], an instantiation of this framework for performing rule updates has been shown to semantically characterise one of the syntax-based rule update semantics. In this paper we show that the framework can also capture a wide range of both model- and formula-based belief update operators which constitute the formal underpinning of existing approaches to ontology updates. Exception-driven operators thus form a unifying perspective on both ontology and rule updates, opening new possibilities for addressing updates of hybrid knowledge bases consisting of both an ontology and a rule component.

Details

ECAI Conference 2010 Conference Paper

On Semantic Update Operators for Answer-Set Programs

Martin Slota
João Leite 0001

Logic programs under the stable models semantics, or answer-set programs, provide an expressive rule based knowledge representation framework, featuring formal, declarative and well-understood semantics. However, handling the evolution of rule bases is still a largely open problem. The AGM framework for belief change was shown to give inappropriate results when directly applied to logic programs under a nonmonotonic semantics such as the stable models. Most approaches to address this issue, developed so far, proposed update operators based on syntactic conditions for rule rejection.

Details

JELIA Conference 2002 Conference Paper

Evolving Logic Programs

José Júlio Alferes
Antonio Brogi
João Leite 0001
Luís Moniz Pereira

Abstract Logic programming has often been considered less than adequate for modelling the dynamics of knowledge changing over time. In this paper we describe a simple though quite powerful approach to modelling the updates of knowledge bases expressed by generalized logic programs, by means of a new language, hereby christened EVOLP (after EVO lving L ogic P rograms). The approach was first sparked by a critical analysis of previous efforts and results in this direction [ 1 ], [ 2 ], [ 7 ], [ 11 ], and aims to provide a simpler, and at once more general, formulation of logic program updating, which runs closer to traditional logic programming (LP) doctrine. From the syntactical point of view, evolving programs are just generalized logic programs (i. e. normal LPs plus default negation also in rule heads), extended with (possibly nested) assertions, whether in heads or bodies of rules. From the semantics viewpoint, a model-theoretic characterization is offered of the possible evolutions of such programs. These evolutions arise both from self (or internal) updating, and from external updating too, originating in the environment. This formulation sets evolving programs on a firm basis in which to express, implement, and reason about dynamic knowledge bases, and opens up a number of interesting research topics that we brush on.

Details

LOPSTR Conference 2001 Conference Paper

Computing Environment-Aware Agent Behaviours with Logic Program Updates

José Júlio Alferes
Antonio Brogi
João Leite 0001
Luís Moniz Pereira

Abstract The ability of reacting to changes in the external environment is of crucial importance within the context of software agents. Such feature must however be suitably reconciled with a more deliberative rational behaviour. In this paper we show how different behaviours of environment-aware agents can be naturally specified and computed in terms of logic program updates. Logic program updates are specified, in our setting, by the language of updates LUPS. We show how such language can be used to model the more reactive behaviours, while keeping the more deliberative features provided by its underlying Dynamic Logic Programming paradigm. The obtained declarative semantics sets a firm basis for the development, implementation, and analysis of environment-aware agents.

Details