Matteo Sammartino

Bio: Matteo Sammartino is an academic researcher from University College London. The author has contributed to research in topics: Automaton & Monad (functional programming). The author has an hindex of 6, co-authored 41 publications receiving 174 citations. Previous affiliations of Matteo Sammartino include Royal Holloway, University of London & Radboud University Nijmegen.

Learning nominal automata

Abstract: We present an Angluin-style algorithm to learn nominal automata, which are acceptors of languages over infinite (structured) alphabets. The abstract approach we take allows us to seamlessly extend known variations of the algorithm to this new setting. In particular we can learn a subclass of nominal non-deterministic automata. An implementation using a recently developed Haskell library for nominal computation is provided for preliminary experiments.

SDN-Actors: Modeling and Verification of SDN Programs

Abstract: Software-Defined Networking (SDN) is a recent networking paradigm that has become increasingly popular in the last decade. It gives unprecedented control over the global behavior of the network and provides a new opportunity for formal methods. Much work has appeared in the last few years on providing bridges between SDN and verification. This paper advances this research line and provides a link between SDN and traditional work on formal methods for verification of distributed software—actor-based modelling. We show how SDN programs can be seamlessly modelled using actors, and thus existing advanced model checking techniques developed for actors can be directly applied to verify a range of properties of SDN networks, including consistency of flow tables, violation of safety policies, and forwarding loops.

CALF: Categorical Automata Learning Framework

Abstract: Automata learning is a technique that has successfully been applied in verification, with the automaton type varying depending on the application domain. Adaptations of automata learning algorithms for increasingly complex types of automata have to be developed from scratch because there was no abstract theory offering guidelines. This makes it hard to devise such algorithms, and it obscures their correctness proofs. We introduce a simple category-theoretic formalism that provides an appropriately abstract foundation for studying automata learning. Furthermore, our framework establishes formal relations between algorithms for learning, testing, and minimization. We illustrate its generality with two examples: deterministic and weighted automata.

The Ï-Calculus: A theory of mobile processes

Abstract: The π-Calculus: A theory of mobile processes by Davide Sangiorgi and David Walker Formal methods have formed the foundation of Computer Science since its inception. Although, initially these formal methods dealt with processes and systems on an individual basis, the paradigm has shifted with the dawn of the age of computer networks. When dealing with systems with interconnected, communicating, dependent, cooperative, and competitive components, the older outlook of analyzing and developing singular systems—and the tools that went with it—were hardly suitable. This led to the development of theories and tools that would support the new paradigm. On the tools end, the development has been widespread and satisfactory: programming languages, development frameworks, databases, and even end-user software products such as word processors, have gained network-awareness. However on the theoretical end, the development has been far less satisfactory. The major work was done by Robin Milner, Joachim Parrow, and David Walker who developed the formalism known as π-calculus in 1989. π-calculus is a process calculus that treats communication between its components as the basic form of computation. It has been quite successful as a foundation of several other calculi in the field and as Milner puts it, it has become common to express ideas about interactions and mobility in variants of the calculus. Introduction The current book serves as a comprehensive reference to π-calculus. Besides Milner's own book on the subject, this is the only other book-length publication on the topic. In many ways, it is much more comprehensive than Milner's: a much wider area of topics are dealt with and in more detail as well. Contents The book is split into seven part. The first part presents the basic theory of π-calculus. However, basic does not mean concise: every topic is discussed in great detail. The section on bisimulation is particularly intensive and provides several insights about the motivation for the theory. Part two discusses several variants of the original calculus. By varying several characteristics of the calculus, such as whether a process can communicate with more than processes at a time, we can obtain these variants. A number of interesting properties of the language are studied by the other when discussing these variants. As can be understood from the title, the calculus' contribution to analyzing mobile processes is a major topic, and it is dealt with extensively starting from part three. The basics are introduced by the way of a sophisticated typing system whose application in speciying complex processes is presented in part four. Part five looks at higher-order π-calculus in which composed systems are considered as first-class citizens. Part six is one of the best in the book and discusses the relation between π-calculus and lambda-calculus, which is an older and more basic calculus. Finally part seven shows how π-calculus can be employed in studying practical, modern software engineering concepts such as object-oriented programming. Impressions One of my disappointments with this book is in how often the reader is left perplexed with some of the theoretical developments, specially in part three. π-calculus is a complicated topic, even for the graduate student audience to which this book is directed, and the author would have done much better by reducing the number of topics and instead focusing on more lucid and detailed explanations. There are several experimental digressions throughout the book, which although interesting, take away from some of the momentum of sequential study. For example, topics such as comparison and encoding of one language to another could be easily moved to a separate section in order to make the content more suitable for self-study. Another issue is the little effort towards making the connection from the theoretical to the practical. The main reason why formal methods have not been adopted in mainstream software development pracitces is that often it is unclear to developers how formalisms can contribute towards the software engineering process. The book would have served its purpose much better if it had spent part of eah chapter discussing the practical application of that chapter's content. For example, congruence checking and bisimulation can be incredbily exciting topics for programmers to learn if they can see practical applications of such powerful techniques. Beyond the above criticism, the book is absolutely indispensible to students and researchers in the field of formal methods. Currently it serves as the primary reference for anyone who wishes to learn the various aspects of π-calculus in detail. Raheel Ahmad

Advances in Model Learning for Software Systems

Abstract: Finding minimal separating sequences for all pairs of inequivalent states in a finite state machine is a classic problem in automata theory. Sets of minimal separating sequences, for instance, play a central role in many conformance testing methods. Moore has already outlined a partition refinement algorithm that constructs such a set of sequences in O(mn) time, where m is the number of transitions and n is the number of states. In this chapter, we present an improved algorithm based on the minimization algorithm of Hopcroft that runs in O(m log n) time. The efficiency of our algorithm is empirically verified and compared to the traditional algorithm.

Learning nominal automata

Abstract: We present an Angluin-style algorithm to learn nominal automata, which are acceptors of languages over infinite (structured) alphabets. The abstract approach we take allows us to seamlessly extend known variations of the algorithm to this new setting. In particular we can learn a subclass of nominal non-deterministic automata. An implementation using a recently developed Haskell library for nominal computation is provided for preliminary experiments.

The power of symbolic automata and transducers

Abstract: Symbolic automata and transducers extend finite automata and transducers by allowing transitions to carry predicates and functions over rich alphabet theories, such as linear arithmetic. Therefore, these models extend their classic counterparts to operate over infinite alphabets, such as the set of rational numbers. Due to their expressiveness, symbolic automata and transducers have been used to verify functional programs operating over lists and trees, to prove the correctness of complex implementations of BASE64 and UTF encoders, and to expose data parallelism in computations that may otherwise seem inherently sequential. In this paper, we give an overview of what is currently known about symbolic automata and transducers as well as their variants. We discuss what makes these models different from their finite-alphabet counterparts, what kind of applications symbolic models can enable, and what challenges arise when reasoning about these formalisms. Finally, we present a list of open problems and research directions that relate to both the theory and practice of symbolic automata and transducers.

Learning Symbolic Automata

Abstract: Symbolic automata allow transitions to carry predicates over rich alphabet theories, such as linear arithmetic, and therefore extend classic automata to operate over infinite alphabets, such as the set of rational numbers. In this paper, we study the foundational problem of learning symbolic automata. We first present \(\mathrm {\Lambda }^*\), a symbolic automata extension of Angluin’s L\(^*\) algorithm for learning regular languages. Then, we define notions of learnability that are parametric in the alphabet theories of the symbolic automata and show how these notions nicely compose. Specifically, we show that if two alphabet theories are learnable, then the theory accepting the Cartesian product or disjoint union of their alphabets is also learnable. Using these properties, we show how existing algorithms for learning automata over large alphabets nicely fall in our framework. Finally, we implement our algorithm in an open-source library and evaluate it on existing automata learning benchmarks.