Model Generation by Moderated Regular Extrapolation

Hagerer, A.; Hungar, Hardi; Niese, Oliver; Steffen, Bernhard

doi:10.1007/3-540-45923-5_6

Cited by 97 publications

(72 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…On Related Work Regular inference techniques have been used for several tasks in verification and test generation, e.g., to create models of environment constraints with respect to which a component should be verified [10], for regression testing to create a specification and test suite [18,21], to perform model checking without access to source code or formal models [16,29], for program analysis [3], and for formal specification and verification [10].…”

Section: Introductionmentioning

confidence: 99%

Learning of Automata Models Extended with Data

Jönsson

2011

Formal Methods for Eternal Networked Software Systems

View full text Add to dashboard Cite

Abstract. One of the challenges in the Connect project is to develop techniques for learning models of networked components from exploratory interaction with the component, based on analyzing messages exchanged between the component and its environment. Many approaches to this problem employ regular inference (aka. automata learning) techniques which generate modest-size finite-state models. Most communication with real-life systems involves data values being relevant to the communication context and thus influencing the observable behavior of the communication endpoints. When applying methods from the realm of automata learning, it is desirable to handle such dataoccurrences. It is therefore important to extend inference techniques to handle message alphabets and state-spaces with structures containing data parameters, often with large domains. After very briefly mentioning several approaches to the problem, we give a longer account of an approach proposed by Aarts et al, which adapts ideas from of predicate abstraction, successfully used in formal verification. We illustrate the techniques by application to a simple running example, which models a simple booking service.Acknowledgment This paper builds on joint work with several present and former collaborators, including Fides Aarts,

show abstract

Section: Introductionmentioning

confidence: 99%

Learning of Automata Models Extended with Data

Jönsson

2011

Formal Methods for Eternal Networked Software Systems

View full text Add to dashboard Cite

show abstract

“…Systems The learning based approaches have fared quite promisingly for the test-based discovery of models of legacy communication systems, thus outperforming prior approaches based on trace combination [3]. As shown in [4,5], the test-based model generation by classical automata learning is very expensive.…”

Section: Communicationmentioning

confidence: 99%

“…[3,20,21] explain our view on the use of learning. Here we only summarize the basic aspects of our realization, which is based on Angluin's learning algorithm L* from [22].…”

Section: Automata Learningmentioning

confidence: 99%

Completing and Adapting Models of Biological Processes

Margaria

Hinchey

Raffelt

et al.

IFIP International Federation for Information Processing

Self Cite

View full text Add to dashboard Cite

Abstract. We present a learning-based method for model completion and adaptation, which is based on the combination of two approaches: 1) R2D2C, a technique for mechanically transforming system requirements via provably equivalent models to running code, and 2) automata learning-bafied model extrapolation. The intended impact of this new combination is to make model completion and adaptation accessible to experts of the field, like biologists or engineers. The principle is briefly illustrated by generating models of biological procedures concerning gene activities in the production of proteins, although the main application is going to concern autonomic systems for space exploration. MotivationA formal approach to Requirements-Based Programming, provisionally named R2D2C ("Requirements to Design to Code"), was developed at NASA [1] as a general-purpose method to mechanically transform system requirements into a provably equivalent model. This is a central need for ultra-high dependability systems like those developed at NASA for space exploration. The R2D2C approach provides mathematically tractable round-trip engineering for system development, rigorously based on formal modelling and formal reasoning techniques. In this paper we complement this method with a learning-based method for model completion and adaptation in order to make model completion and adaptation accessible to experts of the field, like biologists or engineers.Before discussing the technical background and the biological application, we briefly sketch the standard areas of application. Application AreasThe work described below is motivated by the need for requirements-based programming for ultra-high dependability systems which are remote, embedded, and increasingly autonomic.

show abstract

“…This class of techniques has recently started to get attention in the testing and verification community, e.g., for regression testing of telecommunication systems [11,12], and for combining conformance testing and model checking [13,14]. They describe how to construct a finite-state machine (or a regular language) from the answers to a finite sequence of membership queries, each of which observes the component's output in response to a certain input string.…”

Section: Introductionmentioning

confidence: 99%

“…Regular inference techniques have been used for verification and test generation, e.g., to create models of environment constraints with respect to which a component should be verified [19], for regression testing to create a specification and a test suite [11,12], to perform model checking without access to code or to formal models [14,13], for program analysis [20], and for formal specification and verification [19]. Li, Groz, and Shahbaz [21,22] extend regular inference to Mealy machines with a finite subset of input and output symbols from the possible infinite set of symbols.…”

Section: Introductionmentioning

confidence: 99%

Regular Inference for State Machines Using Domains with Equality Tests

Berg

Jönsson

Raffelt

Fundamental Approaches to Software Engineering

View full text Add to dashboard Cite

Abstract. Existing algorithms for regular inference (aka automata learning) allows to infer a finite state machine by observing the output that the machine produces in response to a selected sequence of input strings. We generalize regular inference techniques to infer a class of state machines with an infinite state space. We consider Mealy machines extended with state variables that can assume values from a potentially unbounded domain. These values can be passed as parameters in input and output symbols, and can be used in tests for equality between state variables and/or message parameters. This is to our knowledge the first extension of regular inference to infinite-state systems. We intend to use these techniques to generate models of communication protocols from observations of their input-output behavior. Such protocols often have parameters that represent node adresses, connection identifiers, etc. that have a large domain, and on which test for equality is the only meaningful operation. Our extension consists of two phases. In the first phase we apply an existing inference technique for finite-state Mealy machines to generate a model for the case that the values are taken from a small data domain. In the second phase we transform this finite-state Mealy machine into an infinite-state Mealy machine by folding it into a compact symbolic form.

show abstract

Model Generation by Moderated Regular Extrapolation

Cited by 97 publications

References 17 publications

Learning of Automata Models Extended with Data

Learning of Automata Models Extended with Data

Completing and Adapting Models of Biological Processes

Regular Inference for State Machines Using Domains with Equality Tests

Contact Info

Product

Resources

About