Finding optimal alternatives based on efficient comparative preference inference

Choosing the right or the best option is often a demanding and challenging task for the user (e.g., a customer in an online retailer) when there are many available alternatives. In fact, the user rarely knows which offering will provide the highest value. To reduce the complexity of the choice process, automated recommender systems generate personalized recommendations. These recommendations take into account the preferences collected from the user in an explicit (e.g., letting users express their opinion about items) or implicit (e.g., studying some behavioral features) way. Such systems are widespread; research indicates that they increase the customers' satisfaction and lead to higher sales. Preference handling is one of the core issues in the design of every recommender system. This kind of system often aims at guiding users in a personalized way to interesting or useful options in a large space of possible options. Therefore, it is important for them to catch and model the user's preferences as accurately as possible. In this thesis, we develop a comparative preference-based user model to represent the user's preferences in conversational recommender systems. This type of user model allows the recommender system to capture several preference nuances from the user's feedback. We show that, when applied to conversational recommender systems, the comparative preference-based model is able to guide the user towards the best option while the system is interacting with her. We empirically test and validate the suitability and the practical computational aspects of the comparative preference-based user model and the related preference relations by comparing them to a sum of weights-based user model and the related preference relations. Product configuration, scheduling a meeting and the construction of autonomous agents are among several artificial intelligence tasks that involve a process of constrained optimization, that is, optimization of behavior or options subject to given constraints with regards to a set of preferences. When solving a constrained optimization problem, pruning techniques, such as the branch and bound technique, point at directing the search towards the best assignments, thus allowing the bounding functions to prune more branches in the search tree. Several constrained optimization problems may exhibit dominance relations. These dominance relations can be particularly useful in constrained optimization problems as they can instigate new ways (rules) of pruning non optimal solutions. Such pruning methods can achieve dramatic reductions in the search space while looking for optimal solutions. A number of constrained optimization problems can model the user's preferences using the comparative preferences. In this thesis, we develop a set of pruning rules used in the branch and bound technique to efficiently solve this kind of optimization problem. More specifically, we show how to generate newly defined pruning rules from a dominance algorithm that refers to a set of comparative preferences. These rules include pruning approaches (and combinations of them) which can drastically prune the search space. They mainly reduce the number of (expensive) pairwise comparisons performed during the search while guiding constrained optimization algorithms to find optimal solutions. Our experimental results show that the pruning rules that we have developed and their different combinations have varying impact on the performance of the branch and bound technique.

Leveraging lessons learned in organizations through implementing practice-based organizational learning and performance improvement - An opportunity for context-based intelligent assistant support (CIAS)

Organizations that leverage lessons learned from their experience in the practice of complex real-world activities are faced with five difficult problems. First, how to represent the learning situation in a recognizable way. Second, how to represent what was actually done in terms of repeatable actions. Third, how to assess performance taking account of the particular circumstances. Fourth, how to abstract lessons learned that are re-usable on future occasions. Fifth, how to determine whether to pursue practice maturity or strategic relevance of activities. Here, organizational learning and performance improvement are investigated in a field study using the Context-based Intelligent Assistant Support (CIAS) approach. A new conceptual framework for practice-based organizational learning and performance improvement is presented that supports researchers and practitioners address the problems evoked and contributes to a practice-based approach to activity management. The novelty of the research lies in the simultaneous study of the different levels involved in the activity. Route selection in light rail infrastructure projects involves practices at both the strategic and operational levels; it is part managerial/political and part engineering. Aspectual comparison of practices represented in Contextual Graphs constitutes a new approach to the selection of Key Performance Indicators (KPIs). This approach is free from causality assumptions and forms the basis of a new approach to practice-based organizational learning and performance improvement. The evolution of practices in contextual graphs is shown to be an objective and measurable expression of organizational learning. This diachronic representation is interpreted using a practice-based organizational learning novelty typology. This dissertation shows how lessons learned when effectively leveraged by an organization lead to practice maturity. The practice maturity level of an activity in combination with an assessment of an activity’s strategic relevance can be used by management to prioritize improvement effort.

Caesarean section and subsequent pregnancy outcome: a Danish register-based cohort study

Background and Aims: Caesarean section rates have increased in recent decades and the effects on subsequent pregnancy outcome are largely unknown. Prior research has hypothesised that Caesarean section delivery may lead to an increased risk of subsequent stillbirth, miscarriage, ectopic pregnancy and sub-fertility. Structure and Methods: Papers 1-3 are systematic reviews with meta-analyses. Papers 4-6 are findings from this thesis on the rate of subsequent stillbirth, miscarriage, ectopic pregnancy and live birth by mode of delivery. Results Systematic reviews and meta-analyses: A 23% increased odds of subsequent stillbirth; no increase in odds of subsequent ectopic pregnancy and a 10% reduction in the odds of subsequent live birth among women with a previous Caesarean section were found in the various meta-analyses. Danish cohorts: Results from the Danish Civil Registration System (CRS) cohort revealed a small increased rate of subsequent stillbirth and ectopic pregnancy among women with a primary Caesarean section, which remained in the analyses by type of Caesarean. No increased rate of miscarriage was found among women with a primary Caesarean section. In the CRS data, women with a primary Caesarean section had a significantly reduced rate of subsequent live birth particularly among women with primary elective and maternal-requested Caesarean sections. In the Aarhus Birth Cohort, overall the effect of mode of delivery on the rate and time to next live birth was minimal. Conclusions: Primary Caesarean section was associated with a small increased rate of stillbirth and ectopic pregnancy, which may be in part due to underlying medical conditions. No increased rate of miscarriage was found. A reduced rate of subsequent live birth was found among Caesarean section in the CRS data. In the smaller ABC cohort, a small reduction in rate of subsequent live birth was found among women with a primary Caesarean section and is most likely due to maternal choice rather than any ill effects of the Caesarean. The findings of this study, the largest and most comprehensive to date will be of significant interest to health care providers and women globally.

Synthesis and characterisation of novel superficially porous silica based stationary phases for use in liquid chromatography

The concept of pellicular particles was suggested by Horváth and Lipsky over fifty years ago. The reasoning behind the idea of these particles was to improve column efficiency by shortening the pathways analyte molecules can travel, therefore reducing the effect of the A and C terms. Several types of shell particles were successfully marketed around this time, however with the introduction of high quality fully porous silica under 10 μm, shell particles faded into the background. In recent years a new generation of core shell particles have become popular within the separation science community. These particles allow fast and efficient separations that can be carried out on conventional HPLC systems. Chapter 1 of this thesis introduces the chemistry of chromatographic stationary phases, with an emphasis on silica bonded phases, particularly focusing on the current state of technology in this area. The main focus is on superficially porous silica particles as a support material for liquid chromatography. A summary of the history and development of these particles over the past few decades is explored, along with current methods of synthesis of shell particles. While commercial shell particles have a rough outer surface, Chapter 2 focuses on the novel approach to growth of smooth surface superficially porous particles in a step-by-step manner. From the Stöber methodology to the seeded growth technique, and finally to the layer-bylayer growth of the porous shell. The superficially porous particles generated in this work have an overall diameter of 2.6 μm with a 350 nm porous shell; these silica particles were characterised using SEM, TEM and BET analysis. The uniform spherical nature of the particles along with their surface area, pore size and particle size distribution are examined in this chapter. I discovered that these smooth surface shell particles can be synthesised to give comparable surface area and pore size in comparison to commercial brands. Chapter 3 deals with the bonding of the particles prepared in Chapter 2 with C18 functionality; one with a narrow and one with a wide particle size distribution. This chapter examines the chromatographic and kinetic performance of these silica stationary phases, and compares them to a commercial superficially porous silica phase with a rough outer surface. I found that the particle size distribution does not seem to be the major contributor to the improvement in efficiency. The surface morphology of the particles appears to play an important role in the packing process of these particles and influences the Van Deemter effects. Chapter 4 focuses on the functionalisation of 2.6 μm smooth surface superficially porous particles with a variety of fluorinated and phenyl silanes. The same processes were carried out on 3.0 μm fully porous silica particles to provide a comparison. All phases were accessed using elemental analysis, thermogravimetric analysis, nitrogen sorption analysis and chromatographically evaluated using the Neue test. I observed comparable results for the 2.6 μm shell pentaflurophenyl propyl silica when compared to 3.0 μm fully porous silica. Chapter 5 moves towards nano-particles, with the synthesis of sub-1 μm superficially porous particles, their characterisation and use in chromatography. The particles prepared are 750 nm in total with a 100 nm shell. All reactions and testing carried out on these 750 nm core shell particles are also carried out on 1.5 μm fully porous particles in order to give a comparative result. The 750 nm core shell particles can be synthesised quickly and are very uniform. The main drawback in their use for HPLC is the system itself due to the backpressure experienced using sub – 1 μm particles. The synthesis of modified Stöber particles is also examined in this chapter with a range of non-porous silica and shell silica from 70 nm – 750 nm being tested for use on a Langmuir – Blodgett system. These smooth surface shell particles have only been in existence since 2009. The results displayed in this thesis demonstrate how much potential smooth surface shell particles have provided more in-depth optimisation is carried out. The results on packing studies reported in this thesis aims to be a starting point for a more sophisticated methodology, which in turn can lead to greater chromatographic improvements.

Improved image analysis based corrosion assessment using preprocessing techniques

The importance of non-destructive techniques (NDT) in structural health monitoring programmes is being critically felt in the recent times. The quality of the measured data, often affected by various environmental conditions can be a guiding factor in terms usefulness and prediction efficiencies of the various detection and monitoring methods used in this regard. Often, a preprocessing of the acquired data in relation to the affecting environmental parameters can improve the information quality and lead towards a significantly more efficient and correct prediction process. The improvement can be directly related to the final decision making policy about a structure or a network of structures and is compatible with general probabilistic frameworks of such assessment and decision making programmes. This paper considers a preprocessing technique employed for an image analysis based structural health monitoring methodology to identify sub-marine pitting corrosion in the presence of variable luminosity, contrast and noise affecting the quality of images. A preprocessing of the gray-level threshold of the various images is observed to bring about a significant improvement in terms of damage detection as compared to an automatically computed gray-level threshold. The case dependent adjustments of the threshold enable to obtain the best possible information from an existing image. The corresponding improvements are observed in a qualitative manner in the present study.