Presheaf models for CCS-like languages

An abstract definition of bisimulation is presented. It enables a uniform definition of bisimulation across a range of different models for parallel computation presented as categories. As examples, transition systems, synchronisation trees, transition systems with independence (an abstraction from Petri nets) and labelled event structures are considered. On transition systems the abstract definition readily specialises to Milner's strong bisimulation. On event structures it explains and leads to a revision of history-preserving bisimulation of Rabinovitch and Traktenbrot, Goltz and van Glabeek. A tie-up with open maps in a (pre)topos, as they appear in the work of Joyal and Moerdijk, brings to light a promising new model, presheaves on categories of pomsets, into which the usual category of labelled event structures embeds fully and faithfully. As an indication of its promise, this new presheaf model has ``refinement'' operators, though further work is required to justify their appropriateness and understand their relation to previous attempts. The general approach yields a logic, generalising Hennessy-Milner logic, which is characteristic for the generalised notion of bisimulation.

A systematic treatment of weak bisimulation and observational congruence on presheaf models is presented. The theory is developed with respect to a "hiding" functor from a category of paths to observable paths. Via a view of processes as bundles, we are able to account for weak morphisms (roughly only required to preserve observable paths) and to derive a saturation monad (on the category of presheaves over the category of paths). Weak morphisms may be encoded as strong ones via the Kleisli construction associated to the saturation monad. A general notion of weak open-map bisimulation is introduced, and results relating various notions of strong and weak bisimulation are provided. The abstract theory is accompanied by fine concrete study of two key models for concurrency, the interleaving model of synchronisation trees and the independence model of labelled event structures.

We show how parallel composition of higher-dimensional automata (HDA) can be expressed categorically in the spirit of Winskel & Nielsen. Employing the notion of computation path introduced by van Glabbeek, we define a new notion of bisimulation of HDA using open maps. We derive a connection between computation paths and carrier sequences of dipaths and show that bisimilarity of HDA can be decided by the use of geometric techniques.

What does it mean when we say that computers can ‘write’ and how are recent developments in neural networks and machine learning changing this capacity? This article examines the long-standing literary fear of authorship being replaced by machines while also interrogating the labour and credit implications that sit behind widely used structures of authorship in a technological age. The argument makes reference to one work of computer-generated writing – Johannes Helden & Hakan Jonson’s Evolution [2014] – and to one software paradigm (a character-based recurrent neural networks for language acquisition trained on the corpus of the journal Textual Practice). I here argue that unless we conceive more broadly of the criteria for ‘authorship’ as a labour function, and unless we take seriously the need to see textual production as social production, hybridized (but predominantly) machine identities will come to dominate a literary landscape.

A Fully Abstract Presheaf Semantics of SCCS with Finite Delay

<p>A systematic treatment of weak bisimulation and observational congruence<br />on presheaf models is presented. The theory is developed with<br />respect to a "hiding" functor from a category of paths to observable<br />paths. Via a view of processes as bundles, we are able to account for<br />weak morphisms (roughly only required to preserve observable paths)<br />and to derive a saturation monad (on the category of presheaves over<br />the category of paths). Weak morphisms may be encoded as strong<br />ones via the Kleisli construction associated to the saturation monad.<br />A general notion of weak open-map bisimulation is introduced, and<br />results relating various notions of strong and weak bisimulation are<br />provided. The abstract theory is accompanied by the concrete study<br />of two key models for concurrency, the interleaving model of synchronisation<br /> trees and the independence model of labelled event structures.</p><p>To appear in Proceedings of the 14th Annual IEEE Symposium on Logic<br />in Computer science, LICS'99, IEEE Press, July 1999.</p><p> </p>

This chapter examined the opinion of 417 ESL preservice teachers regarding the use of AI language models in teaching and learning English language. The study found that most of the participants do not believe that AI language models could replace human teachers in ESL education, either now or in the future. However, it was found that there are some ethical, social, and cultural concerns related to the use of AI language models in language education. The concerns voiced by participants include the issue of the users' privacy and ownership, the lack of social interaction, empathy, and communication, and the potential for cultural bias. The chapter also recommended that ESL educators should offer comprehensive training that addresses these concerns and develops critical thinking skills by emphasizing the significance of safeguarding users' privacy and data, mitigating bias in AI models, clarifying intellectual property rights, ensuring reliable text generation, considering social and cultural implications, and promoting a balanced and ethical approach to using AI language models.

Natural language understanding (NLU) and natural language generation (NLG) are two fundamental and related tasks in building task-oriented dialogue systems with opposite objectives: NLU tackles the transformation from natural language to formal representations, whereas NLG does the reverse. A key to success in either task is parallel training data which is expensive to obtain at a large scale. In this work, we propose a generative model which couples NLU and NLG through a shared latent variable. This approach allows us to explore both spaces of natural language and formal representations, and facilitates information sharing through the latent space to eventually benefit NLU and NLG. Our model achieves state-of-the-art performance on two dialogue datasets with both flat and tree-structured formal representations. We also show that the model can be trained in a semi-supervised fashion by utilising unlabelled data to boost its performance.

This paper discusses the principles that should govern the construction of two components of a system for natural language generation (NLG): (1) the ontology - or, rather, as the paper argues, the 'ontological' aspects of a belief system - and (2) the semantic representation of noun senses. It is an interesting fact that many ontologies bear a striking resemblance to a system network, as used in systemic functional grammar (SFG). Furthermore, two major current research efforts in the field of ontology-building are designed to run with a SFG generator: Pangloss, where the generator is Penman, and COMMUNAL, where the generator is GENESYS. It is therefore important to establish a principled approach to the 'division of labour' between the ontology and the equivalent aspects of the model of language - here a system network for the 'meaning potential' of English nouns. (However, the general principles should be relevant to ANY model of language.) The paper summarises (a) the purposes and (b) the structure of (1) a system network for noun senses and (2) the equivalent ontology (based on what we in the COMMUNAL Project judge is required in the next generation of belief systems for NLG). Examples are given of current work on the relevant system network and, more briefly, of the equivalent ontological aspects of the belief system. In particular, reasons are given why it would be inappropriate to give a primary place to the 'mass' vs. 'count' distinction in an 'interlingua' ontology - and even, surprising though it may seem, in a language-specific semantics for English. Finally, it turns out that, in the new perspective presented here, there is no 'component' of the belief system that is 'the ontology', and the reasons for this apparently anomalous position are given.

Probabilistic divide & congruence: Branching bisimilarity

We define a language G for querying data represented as a labeled graph G. By considering G as a relation, this graphical query language can be viewed as a relational query language, and its expressive power can be compared to that of other relational query languages. We do not propose G as an alternative to general purpose relational query languages, but rather as a complementary language in which recursive queries are simple to formulate. The user is aided in this formulation by means of a graphical interface. The provision of regular expressions in G allows recursive queries more general than transitive closure to be posed, although the language is not as powerful as those based on function-free Horn clauses. However, we hope to be able to exploit well-known graph algorithms in evaluating recursive queries efficiently, a topic which has received widespread attention recently.

The article explores the development of language generation technologies, from machine learning models to fluid neural networks for text generation. English is one of the most widespread languages in the world: it is the official language of more than 60 countries. The events of recent years have led to the development of the popularity of the Ukrainian language not only in the country but also abroad. The article analyses scientific sources on this topic, the results of which formed a base for creating a genealogical tree of the development of language synthesis technology. The study pays particular attention to the problem of automating the generation of texts in the Ukrainian language, which includes the absence of strict grammatical constructions of sentences, numerous rules of word formation, presentation of words, placement of accents, and exceptions to almost all rules. The article presents several different transformative architectures that one can use to generate the Ukrainian language. We considered the problems of training such networks and ways to solve them for machine learning models and neural networks. As a result, we determined the stages of development of a neural network for the generation of texts in the Ukrainian language. Even though neural networks are a powerful tool of artificial intelligence, they have certain limitations that the new architecture – liquid neural networks (LNN) – is devoid of. The article also discusses their advantages and disadvantages. The LNN architecture differs from traditional neural networks in its ability to efficiently process continuous or time series data. If new data is available, LNNs can change the number of neurons and connections in a layer. Due to these two main features of LNN, the network expresses the system’s state over time as opposed to the traditional neural network (NN) approach, so it does not require large amounts of observed training data to obtain accurate results. In conclusion, we should note that, despite their shortcomings, LNN networks are more dynamic, adaptive, efficient, and reliable than traditional neural networks and have many potential uses. Keywords: neural network, Ukrainian language, machine learning, speech synthesis, text-to-speech.

The foundation of a process model lies in its control flow specifications. Using a generic process modeling language for workflows, we show how a control flow specification may contain certain structural conflicts that could compromise its correct execution. In general, identification of such conflicts is a computationally complex problem and requires development of effective algorithms specific for target system language. We present a visual verification approach and algorithm that employs a set of graph reduction rules to identify structural conflicts in process models for a generic workflow modeling language. We also provide insights into the correctness and complexity of the reduction process. The main contribution of the paper is a new technique for satisfying well-defined correctness criteria in process models.

Constructing a bidirectional transformation between BPMN and BPEL with a functional logic programming language

This paper deeply studies the characteristics of process language and denotes the process operation language with predicate logic, in the purpose of simplified the generation of process operation language. Using data mining technology, we generalized and stratified the conception of the process operation language. Based on the binding mechanism, a high level model of process operation language was established. The paper has solved the problems in the universal structured generation of the process planning language, and achieved application success.