Computing explanations for interactive constraint-based systems

Constraint programming has emerged as a successful paradigm for modelling combinatorial problems arising from practical situations. In many of those situations, we are not provided with an immutable set of constraints. Instead, a user will modify his requirements, in an interactive fashion, until he is satisfied with a solution. Examples of such applications include, amongst others, model-based diagnosis, expert systems, product configurators. The system he interacts with must be able to assist him by showing the consequences of his requirements. Explanations are the ideal tool for providing this assistance. However, existing notions of explanations fail to provide sufficient information. We define new forms of explanations that aim to be more informative. Even if explanation generation is a very hard task, in the applications we consider, we must manage to provide a satisfactory level of interactivity and, therefore, we cannot afford long computational times. We introduce the concept of representative sets of relaxations, a compact set of relaxations that shows the user at least one way to satisfy each of his requirements and at least one way to relax them, and present an algorithm that efficiently computes such sets. We introduce the concept of most soluble relaxations, maximising the number of products they allow. We present algorithms to compute such relaxations in times compatible with interactivity, achieving this by indifferently making use of different types of compiled representations. We propose to generalise the concept of prime implicates to constraint problems with the concept of domain consequences, and suggest to generate them as a compilation strategy. This sets a new approach in compilation, and allows to address explanation-related queries in an efficient way. We define ordered automata to compactly represent large sets of domain consequences, in an orthogonal way from existing compilation techniques that represent large sets of solutions.

Identifying sources of global contention in constraint satisfaction search

Much work has been done on learning from failure in search to boost solving of combinatorial problems, such as clause-learning and clause-weighting in boolean satisfiability (SAT), nogood and explanation-based learning, and constraint weighting in constraint satisfaction problems (CSPs). Many of the top solvers in SAT use clause learning to good effect. A similar approach (nogood learning) has not had as large an impact in CSPs. Constraint weighting is a less fine-grained approach where the information learnt gives an approximation as to which variables may be the sources of greatest contention. In this work we present two methods for learning from search using restarts, in order to identify these critical variables prior to solving. Both methods are based on the conflict-directed heuristic (weighted-degree heuristic) introduced by Boussemart et al. and are aimed at producing a better-informed version of the heuristic by gathering information through restarting and probing of the search space prior to solving, while minimizing the overhead of these restarts. We further examine the impact of different sampling strategies and different measurements of contention, and assess different restarting strategies for the heuristic. Finally, two applications for constraint weighting are considered in detail: dynamic constraint satisfaction problems and unary resource scheduling problems.

Shining light on food microbiology; applications of luciferase-tagged microorganisms in the food industry

Bioluminescence is the production of light by living organisms as a result of a number of enzyme catalysed reactions caused by enzymes termed luciferases. The lux genes responsible for the emission of light can be cloned from one bioluminescent microorganism into one that is not bioluminescent. The light emitted can be monitored and quantified and will provide information on the metabolic activity, quantity and location of cells in a particular environment, in real-time. The primary aim of this thesis was to investigate and identify several food industry related applications of lux-tagged microorganisms. The first aim was to monitor a lux-tagged Cronobacter sakazakii in reconstituted infant milk formula, in realtime. The second aim was to investigate a bioluminescent-based early warning system for starter culture disruption by bacteriophages and antibiotic residues. The third of this thesis was to examine the use of a bioluminescent-based assay to test the activity of bioengineered Nisin derivatives M21V and S29A against foodborne pathogens in laboratory media and selected foods.

Repairing wireless sensor network connectivity in damaged environments

A wireless sensor network can become partitioned due to node failure, requiring the deployment of additional relay nodes in order to restore network connectivity. This introduces an optimisation problem involving a tradeoff between the number of additional nodes that are required and the costs of moving through the sensor field for the purpose of node placement. This tradeoff is application-dependent, influenced for example by the relative urgency of network restoration. In addition, minimising the number of relay nodes might lead to long routing paths to the sink, which may cause problems of data latency. This data latency is extremely important in wireless sensor network applications such as battlefield surveillance, intrusion detection, disaster rescue, highway traffic coordination, etc. where they must not violate the real-time constraints. Therefore, we also consider the problem of deploying multiple sinks in order to improve the network performance. Previous research has only parts of this problem in isolation, and has not properly considered the problems of moving through a constrained environment or discovering changes to that environment during the repair or network quality after the restoration. In this thesis, we firstly consider a base problem in which we assume the exploration tasks have already been completed, and so our aim is to optimise our use of resources in the static fully observed problem. In the real world, we would not know the radio and physical environments after damage, and this creates a dynamic problem where damage must be discovered. Therefore, we extend to the dynamic problem in which the network repair problem considers both exploration and restoration. We then add a hop-count constraint for network quality in which the desired locations can talk to a sink within a hop count limit after the network is restored. For each new problem of the network repair, we have proposed different solutions (heuristics and/or complete algorithms) which prioritise different objectives. We evaluate our solutions based on simulation, assessing the quality of solutions (node cost, movement cost, computation time, and total restoration time) by varying the problem types and the capability of the agent that makes the repair. We show that the relative importance of the objectives influences the choice of algorithm, and different speeds of movement for the repairing agent have a significant impact on performance, and must be taken into account when selecting the algorithm. In particular, the node-based approaches are the best in the node cost, and the path-based approaches are the best in the mobility cost. For the total restoration time, the node-based approaches are the best with a fast moving agent while the path-based approaches are the best with a slow moving agent. For a medium speed moving agent, the total restoration time of the node-based approaches and that of the path-based approaches are almost balanced.

Reasoning with imprecise trade-offs in decision making under certainty and uncertainty

In many real world situations, we make decisions in the presence of multiple, often conflicting and non-commensurate objectives. The process of optimizing systematically and simultaneously over a set of objective functions is known as multi-objective optimization. In multi-objective optimization, we have a (possibly exponentially large) set of decisions and each decision has a set of alternatives. Each alternative depends on the state of the world, and is evaluated with respect to a number of criteria. In this thesis, we consider the decision making problems in two scenarios. In the first scenario, the current state of the world, under which the decisions are to be made, is known in advance. In the second scenario, the current state of the world is unknown at the time of making decisions. For decision making under certainty, we consider the framework of multiobjective constraint optimization and focus on extending the algorithms to solve these models to the case where there are additional trade-offs. We focus especially on branch-and-bound algorithms that use a mini-buckets algorithm for generating the upper bound at each node of the search tree (in the context of maximizing values of objectives). Since the size of the guiding upper bound sets can become very large during the search, we introduce efficient methods for reducing these sets, yet still maintaining the upper bound property. We define a formalism for imprecise trade-offs, which allows the decision maker during the elicitation stage, to specify a preference for one multi-objective utility vector over another, and use such preferences to infer other preferences. The induced preference relation then is used to eliminate the dominated utility vectors during the computation. For testing the dominance between multi-objective utility vectors, we present three different approaches. The first is based on a linear programming approach, the second is by use of distance-based algorithm (which uses a measure of the distance between a point and a convex cone); the third approach makes use of a matrix multiplication, which results in much faster dominance checks with respect to the preference relation induced by the trade-offs. Furthermore, we show that our trade-offs approach, which is based on a preference inference technique, can also be given an alternative semantics based on the well known Multi-Attribute Utility Theory. Our comprehensive experimental results on common multi-objective constraint optimization benchmarks demonstrate that the proposed enhancements allow the algorithms to scale up to much larger problems than before. For decision making problems under uncertainty, we describe multi-objective influence diagrams, based on a set of p objectives, where utility values are vectors in Rp, and are typically only partially ordered. These can be solved by a variable elimination algorithm, leading to a set of maximal values of expected utility. If the Pareto ordering is used this set can often be prohibitively large. We consider approximate representations of the Pareto set based on ϵ-coverings, allowing much larger problems to be solved. In addition, we define a method for incorporating user trade-offs, which also greatly improves the efficiency.

Context-aware real-time assistant architecture for pervasive healthcare

The aging population in many countries brings into focus rising healthcare costs and pressure on conventional healthcare services. Pervasive healthcare has emerged as a viable solution capable of providing a technology-driven approach to alleviate such problems by allowing healthcare to move from the hospital-centred care to self-care, mobile care, and at-home care. The state-of-the-art studies in this field, however, lack a systematic approach for providing comprehensive pervasive healthcare solutions from data collection to data interpretation and from data analysis to data delivery. In this thesis we introduce a Context-aware Real-time Assistant (CARA) architecture that integrates novel approaches with state-of-the-art technology solutions to provide a full-scale pervasive healthcare solution with the emphasis on context awareness to help maintaining the well-being of elderly people. CARA collects information about and around the individual in a home environment, and enables accurately recognition and continuously monitoring activities of daily living. It employs an innovative reasoning engine to provide accurate real-time interpretation of the context and current situation assessment. Being mindful of the use of the system for sensitive personal applications, CARA includes several mechanisms to make the sophisticated intelligent components as transparent and accountable as possible, it also includes a novel cloud-based component for more effective data analysis. To deliver the automated real-time services, CARA supports interactive video and medical sensor based remote consultation. Our proposal has been validated in three application domains that are rich in pervasive contexts and real-time scenarios: (i) Mobile-based Activity Recognition, (ii) Intelligent Healthcare Decision Support Systems and (iii) Home-based Remote Monitoring Systems.

Real time analysis of atmospheric single particles in urban environments using aerosol time of flight mass spectrometry

In order to determine the size-resolved chemical composition of single particles in real-time an ATOFMS was deployed at urban background sites in Paris and Barcelona during the MEGAPOLI and SAPUSS monitoring campaigns respectively. The particle types detected during MEGAPOLI included several carbonaceous species, metal-containing types and sea-salt. Elemental carbon particle types were highly abundant, with 86% due to fossil fuel combustion and 14% attributed to biomass burning. Furthermore, 79% of the EC was apportioned to local emissions and 21% to continental transport. The carbonaceous particle types were compared with quantitative measurements from other instruments, and while direct correlations using particle counts were poor, scaling of the ATOFMS counts greatly improved the relationship. During SAPUSS carbonaceous species, sea-salt, dust, vegetative debris and various metal-containing particle types were identified. Throughout the campaign the site was influenced by air masses altering the composition of particles detected. During North African air masses the city was heavily influenced by Saharan dust. A regional stagnation was also observed leading to a large increase in carbonaceous particle counts. While the ATOFMS provides a list of particle types present during the measurement campaigns, the data presented is not directly quantitative. The quantitative response of the ATOFMS to metals was examined by comparing the ion signals within particle mass spectra and to hourly mass concentrations of; Na, K, Ca, Ti, V, Cr, Mn, Fe, Zn and Pb. The ATOFMS was found to have varying correlations with these metals depending on sampling issues such as matrix effects. The strongest correlations were observed for Al, Fe, Zn, Mn and Pb. Overall the results of this work highlight the excellent ability of the ATOFMS in providing composition and mixing state information on atmospheric particles at high time resolution. However they also show its limitations in delivering quantitative information directly.

ITSM ProcessGuide – a longitudinal and multi-method field study for real-world DSR artifact evaluation

Process guidance supports users to increase their process model understanding, process execution effectiveness as well as efficiency, and process compliance performance. This paper presents a research in progress encompassing our ongoing DSR project on Process Guidance Systems and a field evaluation of the resulting artifact in cooperation with a company. Building on three theory-grounded design principles, a Process Guidance System artifact for the company’s IT service ticketing process is developed, deployed and used. Fol-lowing a multi-method approach, we plan to evaluate the artifact in a longitudinal field study. Thereby, we will not only gather self-reported but also real usage data. This article describes the development of the artifact and discusses an innovative evaluation approach.

A distributed architecture for the monitoring and analysis of time series data

It is estimated that the quantity of digital data being transferred, processed or stored at any one time currently stands at 4.4 zettabytes (4.4 × 2 70 bytes) and this figure is expected to have grown by a factor of 10 to 44 zettabytes by 2020. Exploiting this data is, and will remain, a significant challenge. At present there is the capacity to store 33% of digital data in existence at any one time; by 2020 this capacity is expected to fall to 15%. These statistics suggest that, in the era of Big Data, the identification of important, exploitable data will need to be done in a timely manner. Systems for the monitoring and analysis of data, e.g. stock markets, smart grids and sensor networks, can be made up of massive numbers of individual components. These components can be geographically distributed yet may interact with one another via continuous data streams, which in turn may affect the state of the sender or receiver. This introduces a dynamic causality, which further complicates the overall system by introducing a temporal constraint that is difficult to accommodate. Practical approaches to realising the system described above have led to a multiplicity of analysis techniques, each of which concentrates on specific characteristics of the system being analysed and treats these characteristics as the dominant component affecting the results being sought. The multiplicity of analysis techniques introduces another layer of heterogeneity, that is heterogeneity of approach, partitioning the field to the extent that results from one domain are difficult to exploit in another. The question is asked can a generic solution for the monitoring and analysis of data that: accommodates temporal constraints; bridges the gap between expert knowledge and raw data; and enables data to be effectively interpreted and exploited in a transparent manner, be identified? The approach proposed in this dissertation acquires, analyses and processes data in a manner that is free of the constraints of any particular analysis technique, while at the same time facilitating these techniques where appropriate. Constraints are applied by defining a workflow based on the production, interpretation and consumption of data. This supports the application of different analysis techniques on the same raw data without the danger of incorporating hidden bias that may exist. To illustrate and to realise this approach a software platform has been created that allows for the transparent analysis of data, combining analysis techniques with a maintainable record of provenance so that independent third party analysis can be applied to verify any derived conclusions. In order to demonstrate these concepts, a complex real world example involving the near real-time capturing and analysis of neurophysiological data from a neonatal intensive care unit (NICU) was chosen. A system was engineered to gather raw data, analyse that data using different analysis techniques, uncover information, incorporate that information into the system and curate the evolution of the discovered knowledge. The application domain was chosen for three reasons: firstly because it is complex and no comprehensive solution exists; secondly, it requires tight interaction with domain experts, thus requiring the handling of subjective knowledge and inference; and thirdly, given the dearth of neurophysiologists, there is a real world need to provide a solution for this domain