A inteligência computacional é considerada por diferentes autores da atualidade como o destino manifesto da Ciência da Computação. A modelagem de diversos aspectos da cognição, tais como aprendizagem e raciocínio, tem sido a motivação para o desenvolvimento dos paradigmas simbólico e conexionista da inteligência artificial e, mais recentemente, para a integração de ambos com o intuito de unificar as vantagens de cada abordagem em um modelo único. Para o desenvolvimento de sistemas inteligentes, bem como para diversas outras áreas da Ciência da Computação, o tempo é considerado como um componente essencial, e a integração de uma dimensão temporal nestes sistemas é fundamental para conseguir uma representação melhor do comportamento cognitivo. Neste trabalho, propomos o SCTL (Sequential Connectionist Temporal Logic), uma abordagem neuro-simbólica para integrar conhecimento temporal, representado na forma de programas em lógica, em redes neurais recorrentes, de forma que a caracterização semântica de ambas representações sejam equivalentes. Além da estratégia para realizar esta conversão entre representações, e da verificação formal da equivalência semântica, também realizamos uma comparação da estratégia proposta com relação a outros sistemas que realizam representação simbólica e temporal em redes neurais. Por outro lado, também descrevemos, de foma algorítmica, o comportamento desejado para as redes neurais geradas, para realizar tanto inferência quanto aprendizagem sob uma ótica temporal. Este comportamento é analisado em diversos experimentos, buscando comprovar o desempenho de nossa abordagem para a modelagem cognitiva considerando diferentes condições e aplicações. / Computational Intelligence is considered, by di erent authors in present days, the manifest destiny of Computer Science. The modelling of di erent aspects of cognition, such as learning and reasoning, has been a motivation for the integrated development of the symbolic and connectionist paradigms of artificial intelligence. More recently, such integration has led to the construction of models catering for integrated learning and reasoning. The integration of a temporal dimension into such systems is a relevant task as it allows for a richer representation of cognitive behaviour features, since time is considered an essential component in intelligent systems development. This work introduces SCTL (Sequential Connectionist Temporal Logic), a neuralsymbolic approach for integrating temporal knowledge, represented as logic programs, into recurrent neural networks. This integration is done in such a way that the semantic characterization of both representations are equivalent. Besides the strategy to achieve translation from one representation to another, and verification of the semantic equivalence, we also compare the proposed approach to other systems that perform symbolic and temporal representation in neural networks. Moreover, we describe the intended behaviour of the generated neural networks, for both temporal inference and learning through an algorithmic approach. Such behaviour is then evaluated by means several experiments, in order to analyse the performance of the model in cognitive modelling under di erent conditions and applications.

A inteligência computacional é considerada por diferentes autores da atualidade como o destino manifesto da Ciência da Computação. A modelagem de diversos aspectos da cognição, tais como aprendizagem e raciocínio, tem sido a motivação para o desenvolvimento dos paradigmas simbólico e conexionista da inteligência artificial e, mais recentemente, para a integração de ambos com o intuito de unificar as vantagens de cada abordagem em um modelo único. Para o desenvolvimento de sistemas inteligentes, bem como para diversas outras áreas da Ciência da Computação, o tempo é considerado como um componente essencial, e a integração de uma dimensão temporal nestes sistemas é fundamental para conseguir uma representação melhor do comportamento cognitivo. Neste trabalho, propomos o SCTL (Sequential Connectionist Temporal Logic), uma abordagem neuro-simbólica para integrar conhecimento temporal, representado na forma de programas em lógica, em redes neurais recorrentes, de forma que a caracterização semântica de ambas representações sejam equivalentes. Além da estratégia para realizar esta conversão entre representações, e da verificação formal da equivalência semântica, também realizamos uma comparação da estratégia proposta com relação a outros sistemas que realizam representação simbólica e temporal em redes neurais. Por outro lado, também descrevemos, de foma algorítmica, o comportamento desejado para as redes neurais geradas, para realizar tanto inferência quanto aprendizagem sob uma ótica temporal. Este comportamento é analisado em diversos experimentos, buscando comprovar o desempenho de nossa abordagem para a modelagem cognitiva considerando diferentes condições e aplicações. / Computational Intelligence is considered, by di erent authors in present days, the manifest destiny of Computer Science. The modelling of di erent aspects of cognition, such as learning and reasoning, has been a motivation for the integrated development of the symbolic and connectionist paradigms of artificial intelligence. More recently, such integration has led to the construction of models catering for integrated learning and reasoning. The integration of a temporal dimension into such systems is a relevant task as it allows for a richer representation of cognitive behaviour features, since time is considered an essential component in intelligent systems development. This work introduces SCTL (Sequential Connectionist Temporal Logic), a neuralsymbolic approach for integrating temporal knowledge, represented as logic programs, into recurrent neural networks. This integration is done in such a way that the semantic characterization of both representations are equivalent. Besides the strategy to achieve translation from one representation to another, and verification of the semantic equivalence, we also compare the proposed approach to other systems that perform symbolic and temporal representation in neural networks. Moreover, we describe the intended behaviour of the generated neural networks, for both temporal inference and learning through an algorithmic approach. Such behaviour is then evaluated by means several experiments, in order to analyse the performance of the model in cognitive modelling under di erent conditions and applications.

In this thesis, we explore the interface between symbolic and dynamical system computation, with particular regard to dynamical system models of neuronal networks. In doing so, we adhere to a definition of computation as the physical realization of a formal system, where we say that a dynamical system performs a computation if a correspondence can be found between its dynamics on a vectorial space and the formal system’s dynamics on a symbolic space. Guided by this definition, we characterize computation in a range of neuronal network models. We first present a constructive mapping between a range of formal systems and Recurrent Neural Networks (RNNs), through the introduction of a Versatile Shift and a modular network architecture supporting its real-time simulation. We then move on to more detailed models of neural dynamics, characterizing the computation performed by networks of delay-pulse-coupled oscillators supporting the emergence of heteroclinic dynamics. We show that a correspondence can be found between these networks and Finite-State Transducers, and use the derived abstraction to investigate how noise affects computation in this class of systems, unveiling a surprising facilitatory effect on information transmission. Finally, we present a new dynamical framework for computation in neuronal networks based on the slow-fast dynamics paradigm, and discuss the consequences of our results for future work, specifically for what concerns the fields of interactive computation and Artificial Intelligence.

A inteligência computacional é considerada por diferentes autores da atualidade como o destino manifesto da Ciência da Computação. A modelagem de diversos aspectos da cognição, tais como aprendizagem e raciocínio, tem sido a motivação para o desenvolvimento dos paradigmas simbólico e conexionista da inteligência artificial e, mais recentemente, para a integração de ambos com o intuito de unificar as vantagens de cada abordagem em um modelo único. Para o desenvolvimento de sistemas inteligentes, bem como para diversas outras áreas da Ciência da Computação, o tempo é considerado como um componente essencial, e a integração de uma dimensão temporal nestes sistemas é fundamental para conseguir uma representação melhor do comportamento cognitivo. Neste trabalho, propomos o SCTL (Sequential Connectionist Temporal Logic), uma abordagem neuro-simbólica para integrar conhecimento temporal, representado na forma de programas em lógica, em redes neurais recorrentes, de forma que a caracterização semântica de ambas representações sejam equivalentes. Além da estratégia para realizar esta conversão entre representações, e da verificação formal da equivalência semântica, também realizamos uma comparação da estratégia proposta com relação a outros sistemas que realizam representação simbólica e temporal em redes neurais. Por outro lado, também descrevemos, de foma algorítmica, o comportamento desejado para as redes neurais geradas, para realizar tanto inferência quanto aprendizagem sob uma ótica temporal. Este comportamento é analisado em diversos experimentos, buscando comprovar o desempenho de nossa abordagem para a modelagem cognitiva considerando diferentes condições e aplicações. / Computational Intelligence is considered, by di erent authors in present days, the manifest destiny of Computer Science. The modelling of di erent aspects of cognition, such as learning and reasoning, has been a motivation for the integrated development of the symbolic and connectionist paradigms of artificial intelligence. More recently, such integration has led to the construction of models catering for integrated learning and reasoning. The integration of a temporal dimension into such systems is a relevant task as it allows for a richer representation of cognitive behaviour features, since time is considered an essential component in intelligent systems development. This work introduces SCTL (Sequential Connectionist Temporal Logic), a neuralsymbolic approach for integrating temporal knowledge, represented as logic programs, into recurrent neural networks. This integration is done in such a way that the semantic characterization of both representations are equivalent. Besides the strategy to achieve translation from one representation to another, and verification of the semantic equivalence, we also compare the proposed approach to other systems that perform symbolic and temporal representation in neural networks. Moreover, we describe the intended behaviour of the generated neural networks, for both temporal inference and learning through an algorithmic approach. Such behaviour is then evaluated by means several experiments, in order to analyse the performance of the model in cognitive modelling under di erent conditions and applications.

No description available.

Ontologies are widely used in various domains, including biomedical research, to structure information, represent knowledge, and analyze data. The combination of ontologies from different domains is crucial for systematic data analysis and comparison of similar domains. This process requires ontology composition, integration, and alignment, which involve creating new classes by reusing classes from different domains, aggregating types of ontologies within the same domain, and finding correspondences between ontologies within the same or similar domain. This thesis presents use cases where we applied ontology composition, integration, and alignment of phenotype ontologies, and evaluated the resulting ontologies and alignment. First, we analyzed a large aging dataset of inbred laboratory mice, using Mouse Anatomy and Mouse Pathology ontologies. Second, we integrated phenotype ontologies for human and model organism phenotypes to enable comparisons of phenotypes between and within individual species. We developed Pheno-e, an extension of PhenomeNet. We identified novel abnormal anatomical classes for fly phenotypes, allowing the annotation of fly genes that were not annotated before. We demonstrate the distinct contributions of each species' phenotypic data to detecting human diseases using Pheno-e, and show that mouse phenotypic data contributes the most to the discovery of gene--disease associations. This work could guide the selection of model organisms when building methods to find gene-disease associations. Additionally, we refined class definitions in phenotypic ontologies, specifically targeting cell cardinality phenotypes. This representation resolved incorrect inferences in the utilized ontologies, enabling accurate interpretation of phenotypic descriptions. Our findings reveal that this correction enhances gene-disease prediction for diseases associated with cardinality phenotypes. Third, we introduce a novel neural-symbolic method that combines logic fundamentals with machine learning for ontology alignment. This method begins with symbolic representation, followed by iterative neural learning for alignment and symbolic representation consistency checking and reasoning, and back to neural learning. We demonstrate that our system generates noncontroversial alignments first and these alignments are coherent with respect to OWL EL. This novel method can pave the way for more accurate and efficient ontology-based methods, which can have significant implications for various semantic web applications.

Artificial neural networks (ANNs) and logic programs have both been suggested as means of modelling human cognition. While ANNs are adaptable and relatively noise resistant, the information they represent is distributed across various neurons and is therefore difficult to interpret. On the contrary, symbolic systems such as logic programs are interpretable but less adaptable. Human cognition is performed in a network of biological neurons and yet is capable of representing symbols, and therefore an ideal model would combine the strengths of the two approaches. This is the goal of Neural-Symbolic Integration [4, 16, 21, 40], in which ANNs are used to produce interpretable, adaptable representations of logic programs and other symbolic models. One neural-symbolic model of reasoning is SHRUTI [89, 95], argued to exhibit biological plausibility in that it captures some aspects of real biological processes. SHRUTI's original developers also suggest that further biological plausibility can be ascribed to the fact that SHRUTI networks can be represented by a model of genetic development [96, 120]. The aims of this thesis are to support the claims of SHRUTI's developers by producing the first such genetic representation for SHRUTI networks and to explore biological plausibility further by investigating the evolvability of the proposed SHRUTI genome. The SHRUTI genome is developed and evolved using principles from Generative and Developmental Systems and Artificial Development [13, 105], in which genomes use indirect encoding to provide a set of instructions for the gradual development of the phenotype just as DNA does for biological organisms. This thesis presents genomes that develop SHRUTI representations of logical relations and episodic facts so that they are able to correctly answer questions on the knowledge they represent. The evolvability of the SHRUTI genomes is limited in that an evolutionary search was able to discover genomes for simple relational structures that did not include conjunction, but could not discover structures that enabled conjunctive relations or episodic facts to be learned. Experiments were performed to understand the SHRUTI fitness landscape and demonstrated that this landscape is unsuitable for navigation using an evolutionary search. Complex SHRUTI structures require that necessary substructures must be discovered in unison and not individually in order to yield a positive change in objective fitness that informs the evolutionary search of their discovery. The requirement for multiple substructures to be in place before fitness can be improved is probably owed to the localist representation of concepts and relations in SHRUTI. Therefore this thesis concludes by making a case for switching to more distributed representations as a possible means of improving evolvability in the future.

In this thesis, we discuss different techniques to bridge the gap between two different approaches to artificial intelligence: the symbolic and the connectionist paradigm. Both approaches have quite contrasting advantages and disadvantages. Research in the area of neural-symbolic integration aims at bridging the gap between them. Starting from a human readable logic program, we construct connectionist systems, which behave equivalently. Afterwards, those systems can be trained, and later the refined knowledge be extracted.

In this thesis, we discuss different techniques to bridge the gap between two different approaches to artificial intelligence: the symbolic and the connectionist paradigm. Both approaches have quite contrasting advantages and disadvantages. Research in the area of neural-symbolic integration aims at bridging the gap between them. Starting from a human readable logic program, we construct connectionist systems, which behave equivalently. Afterwards, those systems can be trained, and later the refined knowledge be extracted.

<p>Language “in the wild” is complex and ambiguous and relies on a shared understanding of the world for its interpretation. Most current natural language processing methods represent language by learning word co-occurrence patterns from massive amounts of linguistic data. This representation can be very powerful, but it is insufficient to capture the meaning behind written and spoken communication. </p> <p> </p> <p>In this dissertation, I will motivate neural-symbolic representations for dealing with these challenges. On the one hand, symbols have inherent explanatory power, and they can help us express contextual knowledge and enforce consistency across different decisions. On the other hand, neural networks allow us to learn expressive distributed representations and make sense of large amounts of linguistic data. I will introduce a holistic framework that covers all stages of the neural-symbolic pipeline: modeling, learning, inference, and its application for diverse discourse scenarios, such as analyzing online discussions, mining argumentative structures, and understanding public discourse at scale. I will show the advantages of neural-symbolic representations with respect to end-to-end neural approaches and traditional statistical relational learning methods.</p> <p><br></p> <p>In addition to this, I will demonstrate the advantages of neural-symbolic representations for learning in low-supervision settings, as well as their capabilities to decompose and explain high-level decision. Lastly, I will explore interactive protocols to help human experts in making sense of large repositories of textual data, and leverage neural-symbolic representations as the interface to inject expert human knowledge in the process of partitioning, classifying and organizing large language resources. </p>