abstract: An old proverb claims that “two heads are better than one”. Crowdsourcing research and practice have taken this to heart, attempting to show that thousands of heads can be even better. This is not limited to leveraging a crowd’s knowledge, but also their creativity—the ability to generate something not only useful, but also novel. In practice, there are initiatives such as Free and Open Source Software communities developing innovative software. In research, the field of crowdsourced creativity, which attempts to design scalable support mechanisms, is blooming. However, both contexts still present many opportunities for advancement. In this dissertation, I seek to advance both the knowledge of limitations in current technologies used in practice as well as the mechanisms that can be used for large-scale support. The overall research question I explore is: “How can we support large-scale creative collaboration in distributed online communities?” I first advance existing support techniques by evaluating the impact of active support in brainstorming performance. Furthermore, I leverage existing theoretical models of individual idea generation as well as recommender system techniques to design CrowdMuse, a novel adaptive large-scale idea generation system. CrowdMuse models users in order to adapt itself to each individual. I evaluate the system’s efficacy through two large-scale studies. I also advance knowledge of current large-scale practices by examining common communication channels under the lens of Creativity Support Tools, yielding a list of creativity bottlenecks brought about by the affordances of these channels. Finally, I connect both ends of this dissertation by deploying CrowdMuse in an Open Source online community for two weeks. I evaluate their usage of the system as well as its perceived benefits and issues compared to traditional communication tools. This dissertation makes the following contributions to the field of large-scale creativity: 1) the design and evaluation of a first-of-its-kind adaptive brainstorming system; 2) the evaluation of the effects of active inspirations compared to simple idea exposure; 3) the development and application of a set of creativity support design heuristics to uncover creativity bottlenecks; and 4) an exploration of large-scale brainstorming systems’ usefulness to online communities. / Dissertation/Thesis / Doctoral Dissertation Computer Science 2019

acase@tulane.edu / 1 / Melissa Schigoda

Innovation has long been considered crucial for companies to gain a competitive edge in the global marketplace. Unfortunately, a solid understanding of the system of innovation does not exist. The literature lacks formal definitions and methodologies for the system of innovation. Many surrogates for innovation metrics have been posited in past research but none have solidified the overall concept of an innovation system or science. It has been speculated that innovation as a system is complex. Additionally, some researchers have suggested that this innovation system is adaptive. In these instances, of the literature, surrogates were again utilized in place of solid modeling and hypothesis that is benchmarked against real world case studies. Surrogates, such as patent citation, do serve a useful purpose to assist in the understanding of the historic nature of the innovation system but they fall short of defining the system completely. This paper seeks to aid in the solidification of a hypothesis of the system of innovation as a complex adaptive system. Initial consideration is directed towards the historic interactions that have taken place in the system of innovation. These interactions are viewed through the surrogate of patent citation as there is little other record of innovation. The novelty of this paper is that patent citations form not the core but rather a starting point for the definition of innovation as a complex adaptive system. Various models are built using techniques of cellular automata as well as agent-based modeling to assist in the understanding of the principles at work in the innovation system. These models present startling evidence that there exists an upper bound on the number of interactions any one invention should utilize in its course towards being deemed an innovation. Additionally, the models describe the benefits of partnership between innovating entities in a rapidly changing marketplace such as the current technological markets. This paper asserts specific conclusions, from the models, that assist in understanding that the system of innovation is truly a complex adaptive system. The models are further supported through real world examples.

One objective of artificial intelligence is to model the behavior of an intelligent agent interacting with its environment. The environment's transformations can be modeled as a Markov chain, whose state is partially observable to the agent and affected by its actions; such processes are known as partially observable Markov decision processes (POMDPs). While the environment's dynamics are assumed to obey certain rules, the agent does not know them and must learn. In this dissertation we focus on the agent's adaptation as captured by the reinforcement learning framework. This means learning a policy---a mapping of observations into actions---based on feedback from the environment. The learning can be viewed as browsing a set of policies while evaluating them by trial through interaction with the environment. The set of policies is constrained by the architecture of the agent's controller. POMDPs require a controller to have a memory. We investigate controllers with memory, including controllers with external memory, finite state controllers and distributed controllers for multi-agent systems. For these various controllers we work out the details of the algorithms which learn by ascending the gradient of expected cumulative reinforcement. Building on statistical learning theory and experiment design theory, a policy evaluation algorithm is developed for the case of experience re-use. We address the question of sufficient experience for uniform convergence of policy evaluation and obtain sample complexity bounds for various estimators. Finally, we demonstrate the performance of the proposed algorithms on several domains, the most complex of which is simulated adaptive packet routing in a telecommunication network.

The complexity of software systems is exploding, along with their use and application in new domains. Managing this complexity has become a focal point for research in Software Engineering. One direction for research in this area is developing techniques for designing adaptive software systems that self-optimize, self-repair, self-configure and self-protect, thereby reducing maintenance costs, while improving quality of service. This thesis presents a requirements-driven approach for developing adaptive and customizable systems. Requirements goal models are used as a basis for capturing problem variability, leading to software designs that support a space of possible behaviours – all delivering the same functionality. This space can be exploited at system deployment time to customize the system on the basis of user preferences. It can also be used at runtime to support system adaptation if the current behaviour of the running system is deemed to be unsatisfactory. The contributions of the thesis include a framework for systematically generating designs from high-variability goal models. Three complementary design views are generated: configurational view (feature model), behavioural view (statecharts) and an architectural view (parameterized architecture). The framework is also applied to the field of business process management for intuitive high-level process customization. In addition, the thesis proposes a modeling framework for capturing domain variability through contexts and applies it to goal models. A single goal model is used to capture requirements variations in different contexts. Models for particular contexts can then be automatically generated from this global requirements model. As well, the thesis proposes a new class of requirements-about-requirements called awareness requirements. Awareness requirements are naturally operationalized through feedback controllers – the core mechanisms of every adaptive system. The thesis presents an approach for systematically designing monitoring, analysis/diagnosis, and compensation components of a feedback controller, given a set of awareness requirements. Situations requiring adaptation are explicitly captured using contexts.

The complexity of software systems is exploding, along with their use and application in new domains. Managing this complexity has become a focal point for research in Software Engineering. One direction for research in this area is developing techniques for designing adaptive software systems that self-optimize, self-repair, self-configure and self-protect, thereby reducing maintenance costs, while improving quality of service. This thesis presents a requirements-driven approach for developing adaptive and customizable systems. Requirements goal models are used as a basis for capturing problem variability, leading to software designs that support a space of possible behaviours – all delivering the same functionality. This space can be exploited at system deployment time to customize the system on the basis of user preferences. It can also be used at runtime to support system adaptation if the current behaviour of the running system is deemed to be unsatisfactory. The contributions of the thesis include a framework for systematically generating designs from high-variability goal models. Three complementary design views are generated: configurational view (feature model), behavioural view (statecharts) and an architectural view (parameterized architecture). The framework is also applied to the field of business process management for intuitive high-level process customization. In addition, the thesis proposes a modeling framework for capturing domain variability through contexts and applies it to goal models. A single goal model is used to capture requirements variations in different contexts. Models for particular contexts can then be automatically generated from this global requirements model. As well, the thesis proposes a new class of requirements-about-requirements called awareness requirements. Awareness requirements are naturally operationalized through feedback controllers – the core mechanisms of every adaptive system. The thesis presents an approach for systematically designing monitoring, analysis/diagnosis, and compensation components of a feedback controller, given a set of awareness requirements. Situations requiring adaptation are explicitly captured using contexts.

The heart and microvasculature have characteristics of a complex adaptive system. Extreme challenges faced by these organ systems cause structural changes which lead to global adaptation. To assess the impact of myocardial interstitial edema on the mechanical properties of the left ventricle and the myocardial interstitium, we induced acute and chronic interstitial edema in dogs. With chronic edema, the primary form of collagen changed from type I to III and left ventricular chamber compliance significantly increased. The resulting functional adaptation allows the chronically edematous heart to maintain left ventricular chamber compliance when challenged with acute edema, thus, preserving cardiac function over a wide range of interstitial fluid pressures. To asses the effect of microvascular occlusions, we reintroduced the Pallid bat wing model and developed a novel mathematical model. We hypothesized that microvessels can switch from predominantly pressure-mediated to shear-mediated responses to ensure dilation during occlusions. Arterioles of unanesthetized Pallid bats were temporarily occluded upstream (n=8) and parallel (n=4) to vessels of interest (20-65 mm). In both cases, the vessels of interest rapidly dilated (36+24 %, 37+33 %), illustrating that they responded appropriately to either decreased pressure or increased shear stress. The model not only reproduced this switching behavior, but reveals its origin as the nonlinear shear-pressure-radius relationship. The properties of the heart and microvasculature were extended to characterize a “Research-Intensive Community” (RIC) model, to provide a feasible solution consistent with the Boyer Commission, to create a sustainable physiology research program. We developed and implemented the model with the aim of aligning diverse goals of participants while simultaneously optimizing research productivity. While the model radically increases the number of undergraduate students supported by a single faculty member, the inherent resilience and scalability of this complex adaptive system enables it to expand without formal institutionalization.

Sustainable management of natural resources requires an in-depth understanding of the interplay between social and ecological change. Linked social-ecological systems (SES) have been described as complex adaptive systems (CAS), which mean that they are irreducible, exhibit nonlinear dynamics, have interactions across scales and are uncertain and unpredictable. These propositions have however rarely been tested empirically, in part due to a lack of methodological approaches and suitable datasets. In this thesis, I address this methodological and empirical gap in a study of long-term change of Baltic Sea fisheries. In Paper I, we develop the concept of fishing style through integrating multivariate statistical analysis and in-depth interviews. We thereby identify an intermediate level of detail for analyzing social-ecological dynamics, embracing the case specific and context dependent approaches of the social sciences with the generalizable and quantifiable approaches from the natural sciences. In Paper II we ask: How has the Baltic Sea fishery been regulated over time, and can we identify a way to quantify regulations in order to be able to analyze their effects? We analyze all regulatory changes in Sweden since 1995 with a new methodology and conclude that there is a clear trend towards increased micro-management. In Paper III, we use the fishing styles developed in Paper I and examine how they have changed over time. We relate these changes to the dynamics of regulation (Paper II), as well as to the dynamics of fish stocks and prices. We conclude that regulation has been the main driving force for observed changes, but also that regulation has prompted significant specialization and decline in flexibility for fishers over time. These changes are unintended consequences and may represent a looming risk for the long-term sustainability of this social-ecological system. Paper IV zooms in on a particular fishery, the pelagic trawl fishery for sprat Sprattus sprattus and Atlantic herring Clupea harengus, mainly targeted for the production of fishmeal and fish oil. Suspicions of non-compliance in this fishery motivated us to apply a statistical approach where we used socioeconomic data to re-estimate the historical catches in this fishery (a novel approach to catch-reconstruction estimates). We found that catches had been significantly underreported over several years, with consequences for the quality of stock assessments and management. The study underlines the importance of understanding linked social, economic and ecological dynamics for sustainable outcomes. Finally, Paper V takes a longer historical look at the Baltic Sea fishery, using regionally disaggregated data from 1914-2009 (96 years), which were analyzed with a novel type of nonlinear statistical time-series methods (Empirical Dynamical Modeling). Our analysis explicitly recognized the potential nonlinear dynamics of SES and showed high predictability across regions of catches and prices of cod Gadus morhua and herring. The signal was generally nonlinear and predictability decreased strongly with time, suggesting that the dynamics of this SES are ever-changing. To our knowledge, this is the first long-term analysis of a SES using empirical data and methods developed from the CAS field of research. The main contributions of this thesis are the integrated analysis of social and ecological data, the development of novel methods for understanding SES dynamics, insights on the ever-changing nature of CAS and the quantitative analysis of management outcomes. Future work should focus on assessing the generality of these findings across a broad range of SES and evaluate alternative governance approaches given the complexity and uncertainty of SES suggested by this thesis. / <p>At the time of the doctoral defense, the following paper was unpublished and had a status as follows: Paper 5: Manuscript.</p>

Today’s software systems are becoming increasingly integrated into the lives of their end-users and their ever-changing environments and needs. These demands lead to a growing complexity of systems. The development of adaptive systems is a promising way to manage this complexity. Adaptive systems are able to adapt their behavior at operation time while considering the changing operational environment to maximize the satisfaction of end-user needs. However, adaptive systems have their own challenges to overcome. Especially, requirements engineering for adaptive systems is challenging given the fact that requirements are active runtime entities and can change at runtime. Requirements engineering activities have not only to take place at design but also at runtime. Requirements engineering for adaptive systems is an emerging research area that has so far received little attention, compared to other research areas (e.g., architecture) for adaptive systems. Adaptive systems need to have a full understanding of the context in order to handle the complexity and satisfy end-user needs. Therefore, a new trend in require- ments engineering for adaptive systems emerged to document requirements with the context in which the requirements are valid. Such contextual requirements necessi- tate adaptive systems to consider and define context in order to fully understand the requirements at operation time. Further, adaptive systems must be able to cope with uncertainty inherent in a changing runtime environment. Otherwise, adaptive sys- tems will not be able to satisfy end-user needs. Therefore, after the system has been deployed, support for the evolution of contextual requirements is needed, too. The trend of considering context as part of a contextual requirement poses new challenges in the field of requirements engineering. This dissertation investigates the capture and evolution of contextual requirements for adaptive systems, which leads to three contributions: First, this dissertation presents a framework that differentiates between context and requirements as two separate entities in contextual requirements that can be captured and can be evolved independently. It is especially necessary to capture and evolve the essential context to support the ability of a system to adapt to fulfilling the needs of its end-users, whose requirements and context are constantly changing. The framework is then applied in two case studies. The first case study investi- gates the usefulness of existing requirements elicitation techniques for the elicitation of contextual requirements. This dissertation’s second contribution is the empirical evidence that existing requirement elicitation techniques can be used for the capture of contextual requirements at design time. We propose a combination of interviews, focus groups and prototyping that we found useful in eliciting contextual requirements in our case study. The second study develops and evaluates techniques to support the evolution of context when contextual requirements are validated at runtime. For this purpose we propose an approach which uses machine learning and feedback loops to support the evolution of contextual requirements and which represents the third contribution of this dissertation. / Graduate

Publish/Subscribe (in short pub/sub) allows clients that share common interest communicate in an asynchronous and loosely-coupled fashion. This paradigm is adopted by many distributed event-driven applications such as social networking services, distributed business processes and cyber-physical systems. These applications cannot afford to have the underlying pub/sub substrate perform unreliably, permanently fail or behave arbitrarily as it will cause significant disturbance to stably serving many end-users. Therefore, a research effort on making pub/sub systems resilient against various failures to sustain high quality of service to the clients is imperative. In this thesis, we focus on the overlay of pub/sub brokers that are widely adopted as a popular architecture for large-scale pub/sub systems. Broker overlays can suffer from various issues such as degradation of topology quality, brokers causing transient or permanent benign failures and Byzantine brokers behaving arbitrarily. We aim to make novel research contributions by exploring fundamental techniques that can help the broker overlays maintain functional and non-functional requirements even under the presence of the aforementioned failures and necessary administrative updates. We first build a set of overlay adaptation primitives that re-configure topologies such as shifting links and replicating brokers. These primitives are designed to involve a small local group of brokers in the pub/sub overlays so that the disruption during the execution of large-scale and dynamic changes can be controlled in a fined-grained manner. For the problem of degrading topology quality, automated planning systems are developed to find a sequence of adaptations that would cause minimal disruption to running services. Also, our primitives can be executed on demand to quickly fail-over a crashed broker or off-load congested brokers. In addition, these on-demand primitives can be used to form a group of dynamically replicated brokers that enforce a novel safety measure to prevent Byzantine brokers from sabotaging the pub/sub overlays. Our contributions are evaluated with systematic consideration of various trade-offs between functional and non-functional properties.