Projecting knowledge graph queries into an embedding space using geometric models (points, boxes and spheres) can help to answer queries for large incomplete knowledge graphs. In this work, we propose a symbolic learning-free approach using fuzzy logic to address the shape-closure problem that restricted geometric-based embedding models to only a few shapes (e.g. ConE) for answering complex logical queries. The use of symbolic approach facilitates non-closure geometric models (e.g. point, box) to handle logical operators (including negation). This enabled our newly proposed spherical embeddings (SpherE) in this work to use a polar coordinate system to effectively represent hierarchical relation. Results show that the SpherE model can answer existential positive first-order logic and negation queries. We show that SpherE significantly outperforms the point and box embeddings approaches while generating semantically meaningful hierarchy-aware embeddings.

Maintenance technicians working in heavy industry play a vital role in sustaining and repairing equipment. Technicians are expected to follow written procedures, usually derived from equipment manuals. Historically maintenance procedures have been paper-based. Organisations are increasingly turning to digital devices, but this transition needs to account for the specific needs of the technicians and the context of their work. In this paper, we assess the impact of a digital tool for maintenance procedures on the work of technicians. To do this, we perform a three-phase industry case study. We use dimensions from the Job Characteristics Model (JCM) (autonomy, feedback, task variety, task identity, task significance) to examine the effects of context-specific design features (i.e., navigation control, task presentation, procedure inputs and feedback, and timed work) on maintenance technicians’ work. Our findings indicate that JCM is a relevant tool for assessing the impact of user interface designs on manual work.

Complex urban systems can be difficult to monitor, diagnose and manage because the complete states of such systems are only partially observable with sensors. State estimation techniques can be used to determine the underlying dynamic behavior of such complex systems with their highly non-linear processes and external time-variant influences. States can be estimated by clustering observed sensor readings. However, clustering performance degrades as the number of sensors and readings (i.e. feature dimension) increases. To address this problem, we propose a framework that learns a feature-centric lower dimensional representation of data for clustering to support analysis of system dynamics. We propose Unsupervised Feature Attention with Compact Representation (UFACR) to rank features contributing to a cluster assignment. These weighted features are then used to learn a reduced-dimension temporal representation of the data with a deep-learning model. The resulting low-dimensional representation can be effectively clustered into states. UFACR is evaluated on real-world and synthetic wastewater treatment plant data sets, and feature ranking outcomes were validated by Wastewater treatment domain experts. Our quantitative and qualitative experimental analyses demonstrate the effectiveness of UFACR for uncovering system dynamics in an automated and unsupervised manner to offer guidance to wastewater engineers to enhance industrial productivity and treatment efficiency.

Maintenance short texts (MST), derived from maintenance work order records, encapsulate crucial information in a concise yet information-rich format. These user-generated technical texts provide critical insights into the state and maintenance activities of machines, infrastructure, and other engineered assets–pillars of the modern economy. Despite their importance for asset management decision-making, extracting and leveraging this information at scale remains a significant challenge. This paper presents MaintIE, a multi-level fine-grained annotation scheme for entity recognition and relation extraction, consisting of 5 top-level classes: PhysicalObject, State, Process, Activity and Property and 224 leaf entities, along with 6 relations tailored to MSTs. Using MaintIE, we have curated a multi-annotator, high-quality, fine-grained corpus of 1,076 annotated texts. Additionally, we present a coarse-grained corpus of 7,000 texts and consider its performance for bootstrapping and enhancing fine-grained information extraction. Using these corpora, we provide model performance measures for benchmarking automated entity recognition and relation extraction. The MaintIE scheme, corpus, and model are publicly available at https://github.com/nlp-tlp/maintie under the MIT license, encouraging further community exploration and innovation in extracting valuable insights from MSTs.

Abstract Long term impacts of shortcuts and compromises taken during software development are described by the metaphor Technical Debt (TD). TD is an emerging business issue in modern interdependent software systems. Our focus is on personnel involved in managing complex adaptive software systems used in automated Remote Operations Centers (ROC’s) responsible for operating industrial equipment more than 1200 km away in the Australian outback. Recognizing and managing TD in these complex adaptive software systems is challenging. This paper builds on advances Serious Games have made toward improving situational awareness, applying these lessons to TD. The prototype Serious Game TD-Sim is tested with employees working in a ROC. Thematic analysis is used to assess pre- and post-game impacts on situational awareness of TD and this is triangulated with in-game information trails. A baseline efficacy of improved TD awareness is established, sufficient to warrant further game development. Results also identify areas for improving future game design. This study is unique, applying a Serious Game to build awareness of TD in complex adaptive software systems, something yet to be covered by traditional TD research.

Deep learning is being widely utilized in industrial process monitoring, control, and optimization. However, in the wastewater industry, its applications are still underexplored. This is because deep learning requires a large amount of labeled training data to induce effective predictive models. Owing to the high cost of sensors and frequency and delay in sampling and laboratory analytics, wastewater treatment process data can be sparse with varying frequencies. One option to address training data limitations is to use transfer learning. However, owing to the large covariate shift between the commonly adopted source domains for transfer learning and the target domain of wastewater processes, this approach leads to unacceptable performance. We address this issue by proposing a novel synthetic data generation method for deep predictive modeling of wastewater plants. Employing a Markov process that utilizes random walk, our technique enables the generation of abundant annotated data for our target domain. The method preserves the temporal dynamics and distribution of the original data, thereby closely mimicking the potential original samples of the domain. We extensively evaluate our method over two different high-rate-algae-based treatment datasets, demonstrating considerable performance gains over existing transfer learning. Our proposed algorithm can assist plant operators to deploy responsive supportive models with limited data.

Recent geometric-based approaches have been shown to efficiently model complex logical queries (including the intersection operation) over Knowledge Graphs based on the natural representation of Venn diagram.Existing geometric-based models (using points, boxes embeddings), however, cannot handle the logical negation operation.Further, those using cones embeddings are limited to representing queries by two-dimensional shapes, which reduced their effectiveness in capturing entities query relations for correct answers.To overcome this challenge, we propose unbounded cylinder embeddings (namely CylE), which is a novel geometric-based model based on threedimensional shapes.Our approach can handle a complete set of basic first-order logic operations (conjunctions, disjunctions and negations).CylE considers queries as Cartesian products of unbounded sector-cylinders and consider a set of nearest boxes corresponds to the set of answer entities.Precisely, the conjunctions can be represented via the intersections of unbounded sector-cylinders.Transforming queries to Disjunctive Normal Form can handle queries with disjunctions.The negations can be represented by considering the closure of complement for an arbitrary unbounded sector-cylinder.Empirical results show that the performance of multihop reasoning task using CylE significantly increases over state-of-the-art geometric-based query embedding models for queries without negation.For queries with negation operations, though the performance is on a par with the best performing geometric-based model, CylE significantly outperforms a recent distributionbased model.

Abstract Group announcement logic (GAL) and coalition announcement logic (CAL) allow us to reason about whether it is possible for groups and coalitions of agents to achieve their desired epistemic goals through truthful public communication. The difference between groups and coalitions in such a context is that the latter make their announcements in the presence of possible adversarial counter-announcements. As epistemic goals may involve some agents remaining ignorant, counter-announcements may preclude coalitions from reaching their goals. We study the relative expressivity of GAL and CAL and provide some results involving their more well-known sibling APAL. We also discuss how the presence of memory alters the relationship between groups and coalition.

In this study we propose obstacle-aware evolution-ary algorithms to identify optimised network topologies for electricity distribution networks including isolated microgrids or stand-alone power systems. We outline the extension of two evo-lutionary algorithms that are modified to consider different types of geographically constrained areas in electricity distribution planning. These areas are represented as polygonal obstacles that either cannot be traversed or cause a higher weight factor when traversing. Both proposed evolutionary algorithms are extended such that they find optimised network solutions that avoid solid obstacles and consider the increased cost of traversing soft obstacles. The algorithms are tested and compared on different types of problem instances with solid and soft obstacles and the problem-specific evolutionary algorithm can be shown to successfully find low cost network topologies on a range of different test instances.

Various extensions of public announcement logic have been proposed with quantification over announcements. The best-known extension is called arbitrary public announcement logic, APAL. It contains a primitive language construct Box phi intuitively expressing that "after every public announcement of a formula, formula phi is true". The logic APAL is undecidable and it has an infinitary axiomatization. Now consider restricting the APAL quantification to public announcements of Boolean formulas only, such that Box phi intuitively expresses that "after every public announcement of a Boolean formula, formula phi is true". This logic can therefore called Boolean arbitrary public announcement logic, BAPAL. The logic BAPAL is the subject of this work. Unlike APAL it has a finitary axiomatization. Also, BAPAL is not at least as expressive as APAL. A further claim that BAPAL is decidable is deferred to a companion paper.

Maintenance of industrial equipment is done by fitters, electricians and other maintainers. For safety and quality control, maintainers must follow procedures; historically these have been paper-based. Asset-owning organisations seek to transition maintainers to digital platforms. However, there are limited studies on the potential impact of digitisation on maintenance work and the maintainers that perform it. Our challenge is to identify interface design considerations that support the safe and reliable execution of work. We looked specifically at maintenance procedures and conducted semi-structured interviews with process-plant maintainers. Thematic analysis identified eight factors influencing maintainers’ perceptions towards using digital technologies in their work. We map these factors to three categories, work identity, agency and community. These categories are consistent with concepts from the Job Characteristics Model (JCM). The contribution of this work is the relevance of job characteristics in guiding user interface design for maintainers, for which we make a number of recommendations.

In this paper we address the Euclidean Steiner tree problem in the plane in the presence of soft and solid polygonal obstacles. The Euclidean Steiner tree problem is a well-known NP-hard problem with different applications in network design. Given a set of terminal nodes in the plane the aim is to find a shortest-length interconnection of the terminals allowing further nodes, so-called Steiner points, to be added. In many real-life scenarios there are further constraints that need to be considered. Regions in the plane that cannot be traversed or can only be traversed at a higher cost can be approximated by polygonal areas that either need to be avoided (solid obstacles) or come with a higher cost of traversing (soft obstacles). We propose a genetic algorithm that uses problem-specific representation and operators to solve this problem and show that the algorithm can solve various test scenarios of different sizes. The presented approach appears to outperform current heuristic approaches for the Steiner tree problem with soft obstacles and was evaluated on larger test instances as well.

The clustering of networks in order to optimise one or more given objectives is a highly researched field with many real-world applications. One of these applications is the clustering of a current or potential future electricity network in order to identify an optimised network topology that could consist of microgrids and stand-alone power systems. This research paper gives a brief overview of the current applications of network partitioning and the different methodologies found in the literature. Then, a novel evolutionary algorithm approach is presented which optimises the topology of rural electricity distribution networks considering a problem-specific objective cost function. Given a set of electricity customer loads and locations, the aim is to identify optimal microgrid and standalone power system formations to minimise the total network costs over a certain time period. The latter part entails a brief introduction to microgrids and some theoretical background, a description of the evaluated cost function, and an outline of the problem-specific evolutionary algorithm used for optimising the network.

Timed automata are powerful models for the analysis of real time systems. The optimal infinite scheduling problem for double-priced timed automata is concerned with finding infinite runs of a system whose long term cost to reward ratio is minimal. Due to the state-space explosion occurring when discretising a timed automaton, exact computation of the optimal infinite ratio is infeasible. This paper describes the implementation and evaluation of ant colony optimisation for approximating the optimal schedule for a given double-priced timed automaton. The application of ant colony optimisation to the corner-point abstraction of the automaton proved generally less effective than a random method. The best found optimisation method was obtained by formulating the choice of time delays in a cycle of the automaton as a linear program and utilizing ant colony optimisation in order to determine a sequence of profitable discrete transitions comprising an infinite behaviour.

It has been shown that every temporal logic formula satisfiable over general linear time has a model than can be expressed as a finite Model Expression (ME). The reals are a subclass of general linear time, so similar techniques can be used for the reals. Although MEs are expressive enough for this task, they represent only a single class of elementary equivalent models. In the case where time is represented by integers, regular expressions are equivalent to automata. An ME is more similar to a single run of an automaton than the automaton itself. In linear time it is often useful to model a system as an automaton (or regular expression) rather than a single run of the automaton. In this paper we extend MEs with the operators from Regular Expressions to produce Regular Model Expressions (RegMEs). It is known that model checking temporal logic formulas over MEs is PSPACE-complete. We show that model checking temporal logic formulas over RegMEs is also PSPACE-complete.

RoCTL* was proposed to model and specify the robustness of reactive systems. RoCTL* extended CTL* with the addition of Obligatory and Robustly operators, which quantify over failure-free paths and paths with one more failure, respectively. This article gives an axiomatization for all the operators of RoCTL* with the exception of the Until operator; this fragment is able to express similar contrary-to-duty obligations to the full RoCTL* logic. We call this formal system NORA, and give a completeness proof. We also consider the fragments of the language containing only path quantifiers but where atoms are dependent on histories. We examine semantic properties and potential axiomatizations for these fragments.

Programmable logic controllers (PLCs) are computers that are hardened for industrial environments and have I/O that are used to monitor and control a physical process. Learning automata specifications from PLCs provides an interface to verification tools that use an automata language, such as Uppaal. This paper introduces the time delay Mealy machine and demonstrates that it is sufficiently expressive to model PLC software. Using the LearnLib library, we implement a custom learning method to learn models from several industrial examples and analyze the efficiency. We show that the method is able to learn from simple PLC software, but the time required to learn increases rapidly with the scale of the software.

Abstract The paper presents a new efficient algorithm for computing worst case execution time (WCET) of systems modelled as timed automata (TA). The algorithm uses a set of abstraction techniques that improve significantly the efficiency of WCET analysis of TA models with cyclic behaviour. We show that the proposed abstractions are exact with respect to the WCET problem in the sense that the WCET computed in the abstract model is equal to the one computed in the concrete model. We also compare our algorithm with the one implemented in the model checker UPPAAL which shows that when infinite cycles exist (i.e. cycles that can be run infinitely often), UPPAAL’s algorithm may not terminate, and when largely repetitive finite cycles exist (i.e. cycles that can be run a large number of times but finite), UPPAAL’s algorithm suffers from the state space explosion, thus leading to a low efficiency or resource exhaustion.

In this paper we investigate the performance of XCSR (a real-valued genetics-based machine learning method) in an online environment in which the feedbacks are received with a delay and not for all the instances. The importance of such environments lies in the fact that many real world environments have these characteristics. For instance, in spam detection some of the undetected spam messages which are delivered to the user may be flagged as spam by user after a while. Hence, the feedback is both delayed and partial in this context. Similar situation can easily be imagined in other fraud detection contexts such as network intrusion and credit card fraud. We also present an architecture for an adaptable online XCSR and present two heuristics to deal with biased partial feedback environments. The heuristics use the information about the environment and their observations and create artificial feedbacks for the classifications that do not receive any feedback. We show that these heuristics always help XCSR learn better and perform more accurately in such situations.

Lauchli and Leonard, in 1966, described a series of operations which are able to build all linear temporal structures up to first order equivalence. More recently these operations have been used to describe executions of continuous systems for the purposes of model checking real-time specifications. In this paper we present an algebra over these operations and show that it is both sound and complete, in that it can generate all equivalences over these models.

Temporal logics over general linear time allow us to capture continuous properties in applications such as distributed systems, natural language, message passing and A.I. modelling of human reasoning. Linear time structures, however, can exhibit a wide range of behaviours that are hard to reason with, or even describe finitely. Recently, a formal language of Model Expressions has been proposed to allow the convenient finite description of an adequately representative range of these generally infinite structures. Given a model described in this Model Expression language and a temporal logic formula, a model checking algorithm decides whether the formula is satisfied at some time in the model. Tools based on such algorithms would support a wide variety of tasks such as verification and counter-example investigation. A previous paper gave an exponential space algorithm for the problem of model checking Until/Since temporal formulas over linear time Model Expressions. Here we prove that the problem is actually PSPACE-complete. We present a new PSPACE algorithm and we show PSPACE-hardness by a reduction from quantified boolean formulas.

We present a formal analysis of the well-known two phase atomic commitment protocol. The protocol is modeled as networks of timed automata using the model checker UPPAAL. The protocol has been verified in two different crash models, the crash-stop model, and the crash-recovery model. The paper also describes how dense-timed model checking technology may be applied to discover the worst case execution time and the corresponding worst-case scenario of the protocol. The analysis also allows us to illustrate various features of the UPPAAL tool, which shows that the specification language of the tool lacks the expressiveness to capture some desired properties of the protocol.

Electronic fraud is highly lucrative, with estimates suggesting these crimes to be worth millions of dollars annually. Because of its complex nature, electronic fraud detection is typically impractical to solve without automation. However, the creation of automated systems to detect fraud is very difficult as adversaries readily adapt and change their fraudulent activities which are often lost in the magnitude of legitimate transactions. This study reviews the most popular types of electronic fraud and the existing nature-inspired detection methods that are used for them. The common characteristics of electronic fraud are examined in detail along with the difficulties and challenges that these present to computational intelligence systems. Finally, open questions and opportunities for further work, including a discussion of emerging types of electronic fraud, are presented to provide a context for ongoing research.

Proving that one language is more succinct than an-other becomes harder when the underlying seman-tics is stronger. We propose to use Formula-Size Games (as put forward by Adler and Immerman, 2003), games that are played on two sets of mod-els, and that directly link the length of play with the size of the formula. Using FSGs, we prove three succinctness results form-dimensional modal logic: (1) In system Km, a notion of ‘everybody knows ’ makes the resulting language exponentially more succinct form> 1 (2) In S5m, the same lan-guage becomes more succinct for m> 3 and (3) Public Announcement Logic is exponentially more succinct than S5m, if m> 3. The latter settles an open problem raised by Lutz, 2006. 1

Historically, learning algorithms have been applied to games as a test of their performance, and with the exponential increases in available computational power, machine learning has been attempted in increasingly complex environments. This paper details the application of neuroevolution of augmenting topologies (NEAT) and accuracy-based learning classifier system (XCS) to the Robocode game environment, with the aim of discovering the ability of each algorithm to learn in this environment. Existing implementations of each algorithm were modified and augmented to allow them to operate in Robocode. In the experiments, performance is measured by pitting the algorithmically-driven players against a series of opponents in various tactical challenges. We conclude by discussing the comparative advantages and disadvantages of both NEAT and XCS as applied to Robocode. Both are able to learn competent strategies, but NEAT is susceptible to overfitting and XCS struggles when many actions are available. The selection of appropriate training scenarios is shown to be a key factor in ensuring that evolved strategies are maximally general for both algorithms.

Fraud detection problems have some uniquely challenging properties which make them difficult. In this paper, we investigate the fraud detection problem by describing the common properties of electronic fraud and examining how learning classifier systems (LCSs) can be applied to it. Also, we introduce "random Boolean function" (RBF); an abstract problem with high level of controllability which can be tuned to exhibit those characteristics individually, and report the results of using XCSR (a continuous variant of LCS) on RBF problem and also on a real-world problem. Results from our experiments demonstrate that XCSR can overcome most of the difficulties inherent to the fraud detection problem and can achieve good performance in case of the real-world problem.

Impact analysis is an activity of assessing the effect of making a set of changes to a software system. Many approaches have been developed include performing impact analysis on a high level model that reflects to low level analysis using class interaction prediction. However, analysis from the model contains false results due to not all interactions between classes have impact to one another. In this paper we introduce a new impact analysis approach that is able to filter some false results using a set of impact prediction filters. The contributions of the paper are: (1) a new impact analysis approach; (2) a new set of impact prediction filters and; (3) evaluation results that show the new impact analysis approach improves the accuracy of the prediction results.

RoCTL* was proposed to model robustness in concurrent systems. RoCTL* extended CTL* with the addition of obligatory and robustly operators, which quantify over failure-free paths and paths with one more failure respectively. Whether RoCTL* is more expressive than CTL* has remained an open problem since the RoCTL* logic was proposed. We use the equivalence of LTL to counter-free automata to show that RoCTL* is expressively equivalent to CTL*; the translation to CTL* provides the first model checking procedure for RoCTL*. However, we show that RoCTL* is relatively succinct as all satisfaction preserving translations into CTL* are non-elementary in length.

This paper presents various semantic interpretations for logics of knowledge and time with prefect recall. We allow both past and future operators and examine the interpretation of different linear flows of time. In particular, we present temporal epistemic logics for each of the following flows of time: arbitrary linear orders; the integers; the rationals; the reals; and for uniform flows of time. (By uniform flows of time, we mean that time is an arbitrary linear order that is common knowledge to all agents). We propose axiomatizations for all logics except the last case, for which we show that no finite axiomatization can be found. The axiomatizations are shown to be sound and complete in the case of arbitrary linear orders and the rationals.

Quantified Propositional Temporal Logic (QPTL) is a linear temporal logic that allows quantification over propositional variables. In the usual semantics for QPTL, a model is an infinite discrete linear sequence of states, with each state having some propositional interpretation. The effect of this is that the interpretation of a proposition at one point in time is independent from its interpretation at another point in time. In this paper, we examine the expressivity and decidability of a QPTL, given generalizations of the usual semantics that do not have this restriction. We introduce the repeating semantics (QPTL/sup R/), which allows states to be repeated throughout a model. While semantic interpretation does not affect the unquantified fragment of QPTL it significantly increases the expressive power in the presence of propositional quantification. In the main result of this paper, we show that QPTL/sup R/ makes the satisfiability problem highly undecidable through a complicated encoding of a tiling problem. We also investigate two less expressive semantics which still allow states to be repeated. We prove the satisfiability problem for one is undecidable, and decidable for the other.