Search

This book constitutes revised selected papers from the thoroughly refereed proceedings of the International Workshop on Coordination, Organizations, Institutions, Norms, and Ethics for Governance of Multi-Agent Systems XIV, COINE 2021, held in London, UK, May 3, 2021. The 9 full papers included in this book were carefully reviewed and selected from 12 submissions. They were organized in topical sections as follows: Invited talk; conceptual frameworks architectures for collaboration and coordination; and modelling and understanding social behaviour using COINE technologies., Part of the book sub series: Lecture Notes in Artificial Intelligence (LNAI)

Background: Fall prevention exercise programmes are known to be effective, but access to these programmes is not always possible. The use of eHealth solutions might be a way forward to increase access and reach a wider population. In this feasibility study the aim was to explore the choice of programme, adherence, and self-reported experiences comparing two exercise programmes – a digital programme and a paper booklet. Methods: A participant preference trial of two self-managed fall prevention exercise interventions. Community-dwelling adults aged 70 years and older exercised independently for four months after one introduction meeting. Baseline information was collected at study start, including a short introduction of the exercise programme, a short physical assessment, and completion of questionnaires. During the four months intervention period, participants self-reported their performed exercises in an exercise diary. At a final meeting, questionnaires about their experiences, and post-assessments, were completed. For adherence analyses data from diaries were used and four subgroups for different levels of participation were compared. Exercise maintenance was followed up with a survey 12 months after study start. Results: Sixty-seven participants, with mean age 77 ± 4 years were included, 72% were women. Forty-three percent chose the digital programme. Attrition rate was 17% in the digital programme group and 37% in the paper booklet group (p = .078). In both groups 50–59% reported exercise at least 75% of the intervention period. The only significant difference for adherence was in the subgroup that completed ≥75% of exercise duration, the digital programme users exercised more minutes per week (p = .001). Participants in both groups were content with their programme but digital programme users reported a significantly higher (p = .026) degree of being content, and feeling supported by the programme (p = .044). At 12 months follow-up 67% of participants using t

Background: Fall prevention exercise programmes are known to be effective, but access to these programmes is not always possible. The use of eHealth solutions might be a way forward to increase access and reach a wider population. In this feasibility study the aim was to explore the choice of programme, adherence, and self-reported experiences comparing two exercise programmes – a digital programme and a paper booklet. Methods: A participant preference trial of two self-managed fall prevention exercise interventions. Community-dwelling adults aged 70 years and older exercised independently for four months after one introduction meeting. Baseline information was collected at study start, including a short introduction of the exercise programme, a short physical assessment, and completion of questionnaires. During the four months intervention period, participants self-reported their performed exercises in an exercise diary. At a final meeting, questionnaires about their experiences, and post-assessments, were completed. For adherence analyses data from diaries were used and four subgroups for different levels of participation were compared. Exercise maintenance was followed up with a survey 12 months after study start. Results: Sixty-seven participants, with mean age 77 ± 4 years were included, 72% were women. Forty-three percent chose the digital programme. Attrition rate was 17% in the digital programme group and 37% in the paper booklet group (p = .078). In both groups 50–59% reported exercise at least 75% of the intervention period. The only significant difference for adherence was in the subgroup that completed ≥75% of exercise duration, the digital programme users exercised more minutes per week (p = .001). Participants in both groups were content with their programme but digital programme users reported a significantly higher (p = .026) degree of being content, and feeling supported by the programme (p = .044). At 12 months follow-up 67% of participants using t

Background: Fall prevention exercise programmes are known to be effective, but access to these programmes is not always possible. The use of eHealth solutions might be a way forward to increase access and reach a wider population. In this feasibility study the aim was to explore the choice of programme, adherence, and self-reported experiences comparing two exercise programmes – a digital programme and a paper booklet. Methods: A participant preference trial of two self-managed fall prevention exercise interventions. Community-dwelling adults aged 70 years and older exercised independently for four months after one introduction meeting. Baseline information was collected at study start, including a short introduction of the exercise programme, a short physical assessment, and completion of questionnaires. During the four months intervention period, participants self-reported their performed exercises in an exercise diary. At a final meeting, questionnaires about their experiences, and post-assessments, were completed. For adherence analyses data from diaries were used and four subgroups for different levels of participation were compared. Exercise maintenance was followed up with a survey 12 months after study start. Results: Sixty-seven participants, with mean age 77 ± 4 years were included, 72% were women. Forty-three percent chose the digital programme. Attrition rate was 17% in the digital programme group and 37% in the paper booklet group (p = .078). In both groups 50–59% reported exercise at least 75% of the intervention period. The only significant difference for adherence was in the subgroup that completed ≥75% of exercise duration, the digital programme users exercised more minutes per week (p = .001). Participants in both groups were content with their programme but digital programme users reported a significantly higher (p = .026) degree of being content, and feeling supported by the programme (p = .044). At 12 months follow-up 67% of participants using t

Background: Fall prevention exercise programmes are known to be effective, but access to these programmes is not always possible. The use of eHealth solutions might be a way forward to increase access and reach a wider population. In this feasibility study the aim was to explore the choice of programme, adherence, and self-reported experiences comparing two exercise programmes – a digital programme and a paper booklet. Methods: A participant preference trial of two self-managed fall prevention exercise interventions. Community-dwelling adults aged 70 years and older exercised independently for four months after one introduction meeting. Baseline information was collected at study start, including a short introduction of the exercise programme, a short physical assessment, and completion of questionnaires. During the four months intervention period, participants self-reported their performed exercises in an exercise diary. At a final meeting, questionnaires about their experiences, and post-assessments, were completed. For adherence analyses data from diaries were used and four subgroups for different levels of participation were compared. Exercise maintenance was followed up with a survey 12 months after study start. Results: Sixty-seven participants, with mean age 77 ± 4 years were included, 72% were women. Forty-three percent chose the digital programme. Attrition rate was 17% in the digital programme group and 37% in the paper booklet group (p = .078). In both groups 50–59% reported exercise at least 75% of the intervention period. The only significant difference for adherence was in the subgroup that completed ≥75% of exercise duration, the digital programme users exercised more minutes per week (p = .001). Participants in both groups were content with their programme but digital programme users reported a significantly higher (p = .026) degree of being content, and feeling supported by the programme (p = .044). At 12 months follow-up 67% of participants using t

Background: Fall prevention exercise programmes are known to be effective, but access to these programmes is not always possible. The use of eHealth solutions might be a way forward to increase access and reach a wider population. In this feasibility study the aim was to explore the choice of programme, adherence, and self-reported experiences comparing two exercise programmes – a digital programme and a paper booklet. Methods: A participant preference trial of two self-managed fall prevention exercise interventions. Community-dwelling adults aged 70 years and older exercised independently for four months after one introduction meeting. Baseline information was collected at study start, including a short introduction of the exercise programme, a short physical assessment, and completion of questionnaires. During the four months intervention period, participants self-reported their performed exercises in an exercise diary. At a final meeting, questionnaires about their experiences, and post-assessments, were completed. For adherence analyses data from diaries were used and four subgroups for different levels of participation were compared. Exercise maintenance was followed up with a survey 12 months after study start. Results: Sixty-seven participants, with mean age 77 ± 4 years were included, 72% were women. Forty-three percent chose the digital programme. Attrition rate was 17% in the digital programme group and 37% in the paper booklet group (p = .078). In both groups 50–59% reported exercise at least 75% of the intervention period. The only significant difference for adherence was in the subgroup that completed ≥75% of exercise duration, the digital programme users exercised more minutes per week (p = .001). Participants in both groups were content with their programme but digital programme users reported a significantly higher (p = .026) degree of being content, and feeling supported by the programme (p = .044). At 12 months follow-up 67% of participants using t

This special issue contains the revised selected papers of ArgLP 2015.

The Eighth International Workshop on Applied Parallel Computing (PARA 2006) was held in Umeå, Sweden, June 18–21, 2006. The workshop was organized by the High Performance Computing Center North (HPC2N) and the Department of Computing Science at Umeå University. The general theme for PARA 2006 was “State of the Art in Scientific and Parallel Computing.” Topics covered at PARA 2006 included basic algorithms and software for scientific, parallel and grid computing, tools and environments for developing high-performance computing applications, as well as a broad spectrum of applications from science and engineering. The workshop included 7 plenary keynote presentations, 15 invited minisymposia organized in 30 sessions, and 16 sessions of contributed talks. The minisymposia and the contributed talks were held in five to six parallel sessions. The main workshop program was preceded by two half-day tutorials. In total, 205 presentations were held at PARA 2006, by speakers representing 28 countries. Extended abstracts for all presentations were made available at the PARA 2006 Web site (www.hpc2n.umu.se/para06). The reviewing process was performed in two stages for evaluation of originality, appropriateness, and significance. In the first stage, extended abstracts were reviewed for selection of contributions to be presented at the workshop. In the second stage the full papers submitted after the workshop were reviewed. In total, 120 papers were selected for publication in this peer-reviewed post-conference proceedings. A number of people contributed in different regards to the organization and the accomplishment of PARA 2006. First of all the Local Organization Committee did a greatly appreciated and enthusiastic job. We also acknowledge the following people for the assistance and support during the workshop days: Yvonne Löwstedt and Anne-Lie Persson; Niklas Edmundsson, Roger Oscarsson, and Mattias Wadenstein. A special thanks goes to the PARA 2006 secretary, Lena Hellman, t

Addressing the management challenges for a multitude of distributed cloud architectures, we focus on the three complementary cloud management problems of predictive elasticity, admission control, and placement (or scheduling) of virtual machines. As these problems are intrinsically intertwined we also propose an approach to optimize the overall system behavior by policy-tuning for the tools handling each of them. Moreover, in order to facilitate the execution of some of the management decisions, we also propose new algorithms for live migration of virtual machines with very high workload and/or over low-bandwidth networks, using techniques such as caching, compression, and prioritization of memory pages., The OPTIMIS project

Anxiety is one of the most critical sources of harm to psychological wellbeing, tied to an array of issues, from discomfort and maladaptive coping to severe pathological disorders -making of anxiety one of the largest economic and social healthcare expenses. AI systems are not neutral to the exposure of individuals and societies to anxiety, and the current emphasis on performance-optimization of current AI systems arguably sets a pathway for a systemic rise of anxiety. As a response to this trend, towards further increasing the human-centeredness of existing applications, this paper is dedicated to depicting the landscape of open challenges, high-impact applications, and promising solutions for designing anxiety-sensitive agents. This paper first circumvents the key components of anxiety through a summary of the extensive psychology literature on anxiety; then shows the feasibility of building agent-based models by putting forward an example of a logical model of anxiety; and last, examines current research fields through the lens of anxiety, highlighting categories of prospective applications and techniques which stand to benefit from anxiety-sensitive agents.

In this modern world of digitalization, abundant amount of data is being generated. This often leads to data of high dimension, making data points far-away from each other. Such data may contain confidential information and must be protected from disclosure. Preserving privacy of this high-dimensional data is still a challenging problem. This paper aims to provide a privacy preserving model to anonymize high-dimensional data maintaining the manifold structure of the data. Manifold Learning hypothesize that real-world data lie on a low-dimensional manifold embedded in a higher-dimensional space. This paper proposes a novel approach that uses geodesic distance in manifold learning methods such as ISOMAP and LLE to preserve the manifold structure on low-dimensional embedding. Later on, anonymization of such sensitive data is achieved by M-MDAV, the manifold version of MDAV using geodesic distance. MDAV is a micro-aggregation privacy model. Finally, to evaluate the efficiency of the prop osed approach machine learning classification is performed on the anonymized lower-embedding. To emphasize the importance of geodesic-manifold learning, we compared our approach with a baseline method in which we try to anonymise high-dimensional data directly without reducing it onto a lower-dimensional space. We evaluate the proposed approach over natural and synthetic data such as tabular, image and textual data sets, and then empirically evaluate the performance of the proposed approach using different evaluation metrics viz. accuracy, precision, recall and K-Stress. We show that our proposed approach is providing accuracy up to 99% and thus, provides a novel contribution of analysing the effects of K-anonymity in manifold learning.

In a previous paper, we proposed an algorithm for ontology-driven conceptual model abstractions [18]. We have implemented and tested this algorithm over a FAIR Catalog of such models represented in the OntoUML language. This provided evidence for the correctness of the algorithm’s implementation, i.e., that it correctly implements the model transformation rules prescribed by the algorithm, and its effectiveness, i.e., it is able to achieve high compression (summarization) rates over these models. However, in addition to these properties, it is fundamental to test the validity of this algorithm, i.e., that it achieves what it is intended to do, namely provide summarizing abstractions over these models whilst preserving the gist of the conceptualization being represented. We performed three user studies to evaluate the usefulness of the resulting abstractions as perceived by modelers. This paper reports on the findings of these user studies and reflects on how they can be exploited to improve the existing algorithm.

In this paper we present necessary properties of logic aggregators and compare their major implementations. If decision making includes the identification of a set of alternatives followed by the evaluation of alternatives and selection of the best alternative, then evaluation must be based on graded logic aggregation. The resulting analytic framework is a graded logic which is a seamless generalization of Boolean logic, based on analytic models of graded simultaneity (various forms of conjunction), graded substitutability (various forms of disjunction) and complementing (negation). These basic logic operations can be implemented in various ways, including means, t-norms/conorms, OWA, and fuzzy integrals. Such mathematical models must be applicable in all regions of the unit hypercube [ 0, 1 ] n. In order to be applicable in various areas of decision making, the logic aggregators must be consistent with observable patterns of human reasoning, supporting both formal logic and semantic aspects of human reasoning. That creates a comprehensive set of logic requirements that logic aggregators must satisfy. Various popular aggregators satisfy these requirements to the extent investigated in this paper. The results of our investigation clearly show the limits of applicability of the analyzed aggregators in the area of decision making.

Player eXperience (PX) testing has attracted attention in the game industry as video games become more complex and widespread. Understanding players’ desires and their experience are key elements to guarantee the success of a game in the highly competitive market. Although a number of techniques have been introduced to measure the emotional aspect of the experience, automated testing of player experience still needs to be explored. This paper presents a framework for automated player experience testing by formulating emotion patterns’ requirements and utilizing a computational model of players’ emotions developed based on a psychological theory of emotions along with a model-based testing approach for test suite generation. We evaluate the strength of our framework by performing mutation test. The paper also evaluates the performance of a search-based generated test suite and LTL model checking-based test suite in revealing various variations of temporal and spatial emotion patterns. Results show the contribution of both algorithms in generating complementary test cases for revealing various emotions in different locations of a game level.

Several high profile incidents that involve Artificial Intelligence (AI) have captured public attention and increased demand for regulation. Low public trust and attitudes towards AI reinforce the need for concrete policy around its development and use. However, current guidelines and standards rolled out by institutions globally are considered by many as high-level and open to interpretation, making them difficult to put into practice. This paper presents ongoing research in the field of Responsible AI and explores numerous methods of operationalising AI ethics. If AI is to be effectively regulated, it must not be considered as a technology alone—AI is embedded in the fabric of our societies and should thus be treated as a socio-technical system, requiring multi-stakeholder involvement and employment of continuous value-based methods of assessment. When putting guidelines and standards into practice, context is of critical importance. The methods and frameworks presented in this paper emphasise this need and pave the way towards operational AI ethics., Part of the book sub series: Lecture Notes in Artificial Intelligence (LNAI)Conference series: ACAI: ECCAI Advanced Course on Artificial Intelligence

In this paper we illustrate a novel architecture generating interpretable behavior and explanations. We refer to this architecture as the Mirror Agent Model because it defines the observer model, that is the target of explicit and implicit communications, as a mirror of the agent's. With the goal of providing a general understanding of this work, we firstly show prior relevant results addressing the informative communication of agents intentions and the production of legible behavior. In the second part of the paper we furnish the architecture with novel capabilities for explanations through off-the-shelf saliency methods, followed by preliminary qualitative results.

Privacy for re-identification, k-anonymity, and differential privacy are the main privacy models considered in the literature on data privacy. We introduced an alternative privacy model called integral privacy, that can be seen as a model for computations avoiding membership inference attacks, as well as other inferences, on aggregates and computations from data (e.g., machine learning models and statistics). In previous papers we have shown how we can compute integrally private statistics (e.g., means and variance), decision trees, and regression. In this paper we introduce clustering with overlapping clusters. The goal is to produce integrally private clusters. We formulate the problem in terms of an optimization problem, and provide a (sub-optimal) solution based on genetic algorithms.

The recently introduced ‘Improved Probabilistic Intuitionistic Fuzzy c-Means algorithm (IPIFCM)’ provides a Probabilistic Euclidean Distance Measure (PEDM) based computationally efficient clustering technique. Since IPIFCM is defined on Type-1 Atanassov Intuitionistic Fuzzy Sets (AIFS), it is unable to capture the uncertainty of the membership and nonmembership values of a given datapoint induced by the hesitancy factor of the AIFS. Interval Type-2 Fuzzy Sets (IT2 FSs) deals with the uncertainty in the membership values. In this paper, we incorporate IT2 FSs in the IPIFCM algorithm by introducing upper bound and lower bound of the membership (and non-membership) values of each datapoint to model the change caused by the hesitancy factor. Accordingly, this paper proposes the ‘Probabilistic Interval Type-2 Intuitionistic clustering algorithm’ (P-IT2IFCM), which uses the interval probabilistic weights for PEDM to propose ‘Interval Type-2 Probabilistic Euclidean Distance Measure’ (IT2PEDM). The proposed algorithm provides superior results to existing Fuzzy c-Means (FCM) algorithms such as the basic FCM algorithm, IFCM algorithm, Kernelized-IFCM algorithm and IPIFCM algorithm when executed over various benchmark UCI datasets.

Anxiety is the most prominent source of stress, harmful behaviours, and psychological disorders. AI systems, usually built for maximizing performance, increase the worldwide exposition to anxiety. This foundational paper introduces Anxiety-Aware Markov Decision Processes (AA-MDPs), the first formalism rooted in fundamental psychology research for modelling the anxiety tied to policies. In addition, this paper formalizes models and practical polynomial algorithms for generating anxiety-sensitive policies. Empirical validation demonstrates that AA-MDPs policies replicate the influence of anxiety on human decision-making observed by fundamental psychology research. Last, this paper demonstrates that AA-MDPs are directly applicable for social good, through a real-world use case (Anxiety-Sensitive Itinerary Planning), the immediate applicability for augmenting any formerly-defined MDP model with anxiety-awareness, and direct tracks developing future high-impact models.

In this paper we illustrate a novel architecture generating interpretable behavior and explanations. We refer to this architecture as the Mirror Agent Model because it defines the observer model, that is the target of explicit and implicit communications, as a mirror of the agent's. With the goal of providing a general understanding of this work, we firstly show prior relevant results addressing the informative communication of agents intentions and the production of legible behavior. In the second part of the paper we furnish the architecture with novel capabilities for explanations through off-the-shelf saliency methods, followed by preliminary qualitative results.

In this paper, we propose two empirical studies to (1) detect Android malware and (2) classify Android malware into families. We first (1) reproduce the results of MalBERT using BERT models learning with Android application’s manifests obtained from 265k applications (vs. 22k for MalBERT) from the AndroZoo dataset in order to detect malware. The results of the MalBERT paper are excellent and hard to believe as a manifest only roughly represents an application, we therefore try to answer the following questions in this paper. Are the experiments from MalBERT reproducible? How important are Permissions for malware detection? Is it possible to keep or improve the results by reducing the size of the manifests? We then (2) investigate if BERT can be used to classify Android malware into families. The results show that BERT can successfully differentiate malware/goodware with 97% accuracy. Furthermore BERT can classify malware families with 93% accuracy. We also demonstrate that Android permissions are not what allows BERT to successfully classify and even that it does not actually need it., Also part of Conference series: ACNS: International Conference on Applied Cryptography and Network Security

Prediction of faults reliably for Air Handling Units (AHU) is a key aspect of correcting errors and eliminating non-optimal functionality. Machine learning classification methods with data sampling are widely utilized to forecast these kinds of events, which, by their nature, seldom occur in equipment. The model proposed in this paper harnesses seven years of data from an air handling unit that contains information about, for example, temperature, humidity, CO2content, and fan speed. This paper contributes to the field of imbalanced classification problems by proposing a novel data undersampling algorithm to enhance the classification model results in the presence of imbalanced and missing data. Moreover, this paper compares several oversampling methods, undersampling methods, probability calibration, and machine learning methods. Then, the paper reports on the proposed final model (proposed undersampling Algorithm 1, Tomek Links, and Logistic Regression) to forecast imperfect heat recovery events in an air handling unit that occur relatively seldom. The precision of the final model was 0.93 for the unseen data; this result was reasonable considering the imbalance of data concurring with missing data sequences.

In this paper we illustrate a novel architecture generating interpretable behavior and explanations. We refer to this architecture as the Mirror Agent Model because it defines the observer model, that is the target of explicit and implicit communications, as a mirror of the agent's. With the goal of providing a general understanding of this work, we firstly show prior relevant results addressing the informative communication of agents intentions and the production of legible behavior. In the second part of the paper we furnish the architecture with novel capabilities for explanations through off-the-shelf saliency methods, followed by preliminary qualitative results.

In this paper we illustrate a novel architecture generating interpretable behavior and explanations. We refer to this architecture as the Mirror Agent Model because it defines the observer model, that is the target of explicit and implicit communications, as a mirror of the agent's. With the goal of providing a general understanding of this work, we firstly show prior relevant results addressing the informative communication of agents intentions and the production of legible behavior. In the second part of the paper we furnish the architecture with novel capabilities for explanations through off-the-shelf saliency methods, followed by preliminary qualitative results.

In this paper we illustrate a novel architecture generating interpretable behavior and explanations. We refer to this architecture as the Mirror Agent Model because it defines the observer model, that is the target of explicit and implicit communications, as a mirror of the agent's. With the goal of providing a general understanding of this work, we firstly show prior relevant results addressing the informative communication of agents intentions and the production of legible behavior. In the second part of the paper we furnish the architecture with novel capabilities for explanations through off-the-shelf saliency methods, followed by preliminary qualitative results.

In a recent paper we introduced differentially private random dictatorship as a private mechanism for social choice. Differentially private mechanisms are evaluated in terms of their utility and information loss. In the area of social choice it is not so straightforward to evaluate the utility of a mechanism. It is therefore difficult to evaluate a differentially private social choice mechanism. In this paper we propose to use a particle swarm optimization-like problem to evaluate our differentially private social choice method. Standard particle swarm optimization (PSO) can be seen in terms of a panopticon structure. That is, a structure in which there is a central entity that knows all of all. In PSO, there is a particle or agent that knows the best position achieved by any of the particles or agents. We propose here PSO without panopticon as a way to avoid an omniscient agent in the PSO system. Then, we compare different social choice mechanisms for this PSO without panopticon, and we show that differentially private random dictatorship leads to good results.

This paper introduces an agent framework that integrates the Belief, Desire, Intention (BDI) model with Multi-Context Systems (MCS), particularly for dealing with diverse knowledge sources in belief revision, deliberation and means-end reasoning. By specifying a separate MCS in each BDI-component, the framework manages the interaction between, possibly conflicting, sets of beliefs, desires, intentions and plans generated by specialized sub-systems. A MCS-based BDI-component generates an equilibrium. An approach is introduced for transferring equilibrium between MCSs according to the BDI control loop. This involves the translation of knowledge bases to Answer Set Programming (ASP) to build a shared logic. The proposed framework contributes to the advancement of hybrid intentional agents, where multiple goals and plans must be interwoven in order to deal with a complex multi-modal domain. The potential of the framework is illustrated in a running example, where a driving assistant agent is designed to manage diverse mental states of a human driver, such as emotions, motivations and norms, within each stage of the BDI control loop, producing a plan that is in balance with the diverse contexts.

This paper presents BPMN2Constraints, a tool that compiles BPMN diagrams into sets of declarative con-straints that can then, for example, be used for conformance checking. Notably, BPMN2Constraints doesnot rely on Petri net replay for generating the constraints; by generating constraints directly from a con-trol flow graph extracted from the BPMN model, the tool avoids indirection. BPMN2Constraints cangenerate constraints in several languages: DECLARE, finite-trace linear temporal logic, and SIGNAL, aproprietary process querying language.

The field of Virtual Knowledge Graphs (VKGs) continues to grow in both academic and applied contexts. Yet, the issue of updates in VKG systems has not yet received adequate attention, although it is crucial to manage data modifications at the data source level through the lens of an ontology. In this paper, we focus on VKGs whose ontology is specified in the lightweight ontology language DL-LiteA, and we propose diverse settings and research directions we intend to explore to address the challenge of translating ontology-based updates into updates at the level of data sources. We also pay attention to the important problem of automated analysis of mappings, which plays a major role when it comes to reformulating ontology-based update requests into update requests over the data sources.

In this paper, we formalize practical byte pair encoding tokenization as it is used in large language models and other NLP systems, in particular we formally define and investigate the semantics of the SentencePiece and HuggingFace tokenizers, in particular how they relate to each other, depending on how the tokenization rules are constructed. Beyond this we consider how tokenization can be performed in an incremental fashion, as well as doing it left-to-right using an amount of memory constant in the length of the string, enabling e.g. using a finite state string-to-string transducer.

In comparison to computational linguistics, with its abundance of natural-language datasets, corpora of music analyses are rather fewer and generally smaller. This is partly due to difficulties inherent to the encoding of music analyses, whose multimodal representations—typically a combination of music notation, graphic notation, and natural language—are designed for communication between human musician-analysts, not for automated large-scale data analysis. Analyses based on hierarchical models of tonal structure, such as Heinrich Schenker’s, present additional notational and encoding challenges, since they establish relations between non- adjacent tones, and typically interpret successions of tones as expressions of abstract chordal sonorities, which may not be literally present in the music score. Building on a published XML format by Rizo and Marsden (2019), which stores analyses alongside symbolically encoded scores, this paper presents a generic graph model for reasoning about music analyses, as well as a graphical web application for creating and encoding music analyses in the aforementioned XML format. Several examples are given showing how various techniques of music analysis, primarily but not necessarily hierarchical, might be unambiguously represented through this model.

By identifying and characterising the narratives told in news media we can better understand political and societal processes. The problem is challenging from the perspective of natural language processing because it requires a combination of quantitative and qualitative methods. This paper reports on work in progress, which aims to build a human-in-the-loop pipeline for analysing how the variation of narrative themes across different domains, based on topic modelling and word embeddings. As an illustration, we study the language associated with the threat narrative in British news media.

One important aspect when designing understandable robots is how robots should communicate with a human user to be understood in the best way. In elder care applications this is particularly important, and also difficult since many older adults suffer from various kinds of impairments. In this paper we present a solution where communication modality and communication parameters are adapted to fit both a user profile and an environment model comprising information about light and sound conditions that may affect communication. The Rasa dialogue manager is complemented with necessary functionality, and the operation is verified with a Pepper robot interacting with several personas with impaired vision, hearing, and cognition. Several relevant ethical questions are identified and briefly discussed, as a contribution to the WARN workshop.

The complexity and relative expressiveness of Ontology-mediated Queries (OMQs) is quite well understood by now. In this paper, we study the expressive power of OMQs from a descriptive complexity perspective, where the central question is to understand whether a given OMQ language is powerful enough to express all queries that can be computed within some bound on time or space. We show that the OMQ language that pairs instance queries with ontologies in the very expressive DL ALCHOI with closed predicates cannot express all coNP-computable Boolean queries, despite being coNP-complete in data complexity. We, then, propose an extension of this OMQ language that is expressive enough to precisely capture the class of all Boolean queries computable in coNP. This involves adding functionality as well as path expressions and nominal schemata, which are restricted in a way that allows us to carefully incorporate them into the existing mosaic technique for the DL ALCHOIF with closed predicates without affecting the coNP upper bound in data complexity.

Circumscription is a prominent approach to bring non-monotonicity to Description Logics (DLs), but unfortunately, it usually displays very high computational complexity of reasoning. Many works have studied circumscribed DLs, but most of them focus on expressive DLs containing ALC, and the results for low-complexity DLs are limited. This paper summarises some recent progress in characterizing the computational complexity of reasoning in circumscribed DL-Lite. We perform a two-dimensional analysis, considering different languages of the DL-Lite family, and varying how concepts and roles are treated. In addition to classical circumscription, we consider the recently studied pointwise circumscription, which shows better complexity, in some cases, and remains decidable in the presence of minimized roles.

The development in smart meter technology has made grid operations more efficient based on fine-grained electricity usage data generated at different levels of time granularity. Consequently, machine learning algorithms have benefited from these data to produce useful models for important grid operations. Although machine learning algorithms need historical data to improve predictive performance, these data are not readily available for public utilization due to privacy issues. The existing smart grid data simulation frameworks generate grid data with implicit privacy concerns since the data are simulated from a few real energy consumptions that are publicly available. This paper addresses two issues in smart grid. First, it assesses the level of privacy violation with the individual household appliances based on synthetic household aggregate loads consumption. Second, based on the findings, it proposes two privacy-preserving mechanisms to reduce this risk. Three inference attacks are simulated and the results obtained confirm the efficacy of the proposed privacy-preserving mechanisms.

This paper introduces an agent framework that integrates the Belief, Desire, Intention (BDI) model with Multi-Context Systems (MCS), particularly for dealing with diverse knowledge sources in belief revision, deliberation and means-end reasoning. By specifying a separate MCS in each BDI-component, the framework manages the interaction between, possibly conflicting, sets of beliefs, desires, intentions and plans generated by specialized sub-systems. A MCS-based BDI-component generates an equilibrium. An approach is introduced for transferring equilibrium between MCSs according to the BDI control loop. This involves the translation of knowledge bases to Answer Set Programming (ASP) to build a shared logic. The proposed framework contributes to the advancement of hybrid intentional agents, where multiple goals and plans must be interwoven in order to deal with a complex multi-modal domain. The potential of the framework is illustrated in a running example, where a driving assistant agent is designed to manage diverse mental states of a human driver, such as emotions, motivations and norms, within each stage of the BDI control loop, producing a plan that is in balance with the diverse contexts.

Avoiding bias and understanding the consequences of artificial intelligence used in decision making is of high importance to avoid mistreatment and unintended harm. This paper aims to present an impact focused approach to model the information flow of a socio-technical decision system for analysis of bias and fairness. The framework roots otherwise abstract technical accuracy and bias measures in stakeholder effects and forms a scaffold around which further analysis of the socio-technical system and its components can be coordinated. Two example use-cases are presented and analysed.

The field of Virtual Knowledge Graphs (VKGs) continues to grow in both academic and applied contexts. Yet, the issue of updates in VKG systems has not yet received adequate attention, although it is crucial to manage data modifications at the data source level through the lens of an ontology. In this paper, we focus on VKGs whose ontology is specified in the lightweight ontology language DL-LiteA, and we propose diverse settings and research directions we intend to explore to address the challenge of translating ontology-based updates into updates at the level of data sources. We also pay attention to the important problem of automated analysis of mappings, which plays a major role when it comes to reformulating ontology-based update requests into update requests over the data sources.

In this paper, we formalize practical byte pair encoding tokenization as it is used in large language models and other NLP systems, in particular we formally define and investigate the semantics of the SentencePiece and HuggingFace tokenizers, in particular how they relate to each other, depending on how the tokenization rules are constructed. Beyond this we consider how tokenization can be performed in an incremental fashion, as well as doing it left-to-right using an amount of memory constant in the length of the string, enabling e.g. using a finite state string-to-string transducer.

In comparison to computational linguistics, with its abundance of natural-language datasets, corpora of music analyses are rather fewer and generally smaller. This is partly due to difficulties inherent to the encoding of music analyses, whose multimodal representations—typically a combination of music notation, graphic notation, and natural language—are designed for communication between human musician-analysts, not for automated large-scale data analysis. Analyses based on hierarchical models of tonal structure, such as Heinrich Schenker’s, present additional notational and encoding challenges, since they establish relations between non- adjacent tones, and typically interpret successions of tones as expressions of abstract chordal sonorities, which may not be literally present in the music score. Building on a published XML format by Rizo and Marsden (2019), which stores analyses alongside symbolically encoded scores, this paper presents a generic graph model for reasoning about music analyses, as well as a graphical web application for creating and encoding music analyses in the aforementioned XML format. Several examples are given showing how various techniques of music analysis, primarily but not necessarily hierarchical, might be unambiguously represented through this model.

The development in smart meter technology has made grid operations more efficient based on fine-grained electricity usage data generated at different levels of time granularity. Consequently, machine learning algorithms have benefited from these data to produce useful models for important grid operations. Although machine learning algorithms need historical data to improve predictive performance, these data are not readily available for public utilization due to privacy issues. The existing smart grid data simulation frameworks generate grid data with implicit privacy concerns since the data are simulated from a few real energy consumptions that are publicly available. This paper addresses two issues in smart grid. First, it assesses the level of privacy violation with the individual household appliances based on synthetic household aggregate loads consumption. Second, based on the findings, it proposes two privacy-preserving mechanisms to reduce this risk. Three inference attacks are simulated and the results obtained confirm the efficacy of the proposed privacy-preserving mechanisms.

Circumscription is a prominent approach to bring non-monotonicity to Description Logics (DLs), but unfortunately, it usually displays very high computational complexity of reasoning. Many works have studied circumscribed DLs, but most of them focus on expressive DLs containing ALC, and the results for low-complexity DLs are limited. This paper summarises some recent progress in characterizing the computational complexity of reasoning in circumscribed DL-Lite. We perform a two-dimensional analysis, considering different languages of the DL-Lite family, and varying how concepts and roles are treated. In addition to classical circumscription, we consider the recently studied pointwise circumscription, which shows better complexity, in some cases, and remains decidable in the presence of minimized roles.

The complexity and relative expressiveness of Ontology-mediated Queries (OMQs) is quite well understood by now. In this paper, we study the expressive power of OMQs from a descriptive complexity perspective, where the central question is to understand whether a given OMQ language is powerful enough to express all queries that can be computed within some bound on time or space. We show that the OMQ language that pairs instance queries with ontologies in the very expressive DL ALCHOI with closed predicates cannot express all coNP-computable Boolean queries, despite being coNP-complete in data complexity. We, then, propose an extension of this OMQ language that is expressive enough to precisely capture the class of all Boolean queries computable in coNP. This involves adding functionality as well as path expressions and nominal schemata, which are restricted in a way that allows us to carefully incorporate them into the existing mosaic technique for the DL ALCHOIF with closed predicates without affecting the coNP upper bound in data complexity.

Avoiding bias and understanding the consequences of artificial intelligence used in decision making is of high importance to avoid mistreatment and unintended harm. This paper aims to present an impact focused approach to model the information flow of a socio-technical decision system for analysis of bias and fairness. The framework roots otherwise abstract technical accuracy and bias measures in stakeholder effects and forms a scaffold around which further analysis of the socio-technical system and its components can be coordinated. Two example use-cases are presented and analysed.

This paper introduces an agent framework that integrates the Belief, Desire, Intention (BDI) model with Multi-Context Systems (MCS), particularly for dealing with diverse knowledge sources in belief revision, deliberation and means-end reasoning. By specifying a separate MCS in each BDI-component, the framework manages the interaction between, possibly conflicting, sets of beliefs, desires, intentions and plans generated by specialized sub-systems. A MCS-based BDI-component generates an equilibrium. An approach is introduced for transferring equilibrium between MCSs according to the BDI control loop. This involves the translation of knowledge bases to Answer Set Programming (ASP) to build a shared logic. The proposed framework contributes to the advancement of hybrid intentional agents, where multiple goals and plans must be interwoven in order to deal with a complex multi-modal domain. The potential of the framework is illustrated in a running example, where a driving assistant agent is designed to manage diverse mental states of a human driver, such as emotions, motivations and norms, within each stage of the BDI control loop, producing a plan that is in balance with the diverse contexts.

In this paper, we formalize practical byte pair encoding tokenization as it is used in large language models and other NLP systems, in particular we formally define and investigate the semantics of the SentencePiece and HuggingFace tokenizers, in particular how they relate to each other, depending on how the tokenization rules are constructed. Beyond this we consider how tokenization can be performed in an incremental fashion, as well as doing it left-to-right using an amount of memory constant in the length of the string, enabling e.g. using a finite state string-to-string transducer.

In comparison to computational linguistics, with its abundance of natural-language datasets, corpora of music analyses are rather fewer and generally smaller. This is partly due to difficulties inherent to the encoding of music analyses, whose multimodal representations—typically a combination of music notation, graphic notation, and natural language—are designed for communication between human musician-analysts, not for automated large-scale data analysis. Analyses based on hierarchical models of tonal structure, such as Heinrich Schenker’s, present additional notational and encoding challenges, since they establish relations between non- adjacent tones, and typically interpret successions of tones as expressions of abstract chordal sonorities, which may not be literally present in the music score. Building on a published XML format by Rizo and Marsden (2019), which stores analyses alongside symbolically encoded scores, this paper presents a generic graph model for reasoning about music analyses, as well as a graphical web application for creating and encoding music analyses in the aforementioned XML format. Several examples are given showing how various techniques of music analysis, primarily but not necessarily hierarchical, might be unambiguously represented through this model.

The field of Virtual Knowledge Graphs (VKGs) continues to grow in both academic and applied contexts. Yet, the issue of updates in VKG systems has not yet received adequate attention, although it is crucial to manage data modifications at the data source level through the lens of an ontology. In this paper, we focus on VKGs whose ontology is specified in the lightweight ontology language DL-LiteA, and we propose diverse settings and research directions we intend to explore to address the challenge of translating ontology-based updates into updates at the level of data sources. We also pay attention to the important problem of automated analysis of mappings, which plays a major role when it comes to reformulating ontology-based update requests into update requests over the data sources.