Modeling the Network of Dutch and Flemish Print Production, 1550-1750

The production of artistic prints in the sixteenth- and seventeenth-century Netherlands was an inherently social process. Turning out prints at any reasonable scale depended on the fluid coordination between designers, platecutters, and publishers; roles that, by the sixteenth century, were considered distinguished enough to merit distinct credits engraved on the plates themselves: invenit, fecit/sculpsit, and excudit. While any one designer, plate cutter, and publisher could potentially exercise a great deal of influence over the production of a single print, their individual decisions (Whom to select as an engraver? What subjects to create for a print design? What market to sell to?) would have been variously constrained or encouraged by their position in this larger network (Who do they already know? And who, in turn, do their contacts know?) This dissertation addresses the impact of these constraints and affordances through the novel application of computational social network analysis to major databases of surviving prints from this period. This approach is used to evaluate several questions about trends in early modern print production practices that have not been satisfactorily addressed by traditional literature based on case studies alone: Did the social capital demanded by print production result in centralized, or distributed production of prints? When, and to what extent, did printmakers and publishers in the Low countries favor international versus domestic collaborators? And were printmakers under the same pressure as painters to specialize in particular artistic genres? This dissertation ultimately suggests how simple professional incentives endemic to the practice of printmaking may, at large scales, have resulted in quite complex patterns of collaboration and production. The framework of network analysis surfaces the role of certain printmakers who tend to be neglected in aesthetically-focused histories of art. This approach also highlights important issues concerning art historians’ balancing of individual influence versus the impact of longue durée trends. Finally, this dissertation also raises questions about the current limitations and future possibilities of combining computational methods with cultural heritage datasets in the pursuit of historical research.

Learning Binary Code Representations for Effective and Efficient Image Retrieval

The size of online image datasets is constantly increasing. Considering an image dataset with millions of images, image retrieval becomes a seemingly intractable problem for exhaustive similarity search algorithms. Hashing methods, which encodes high-dimensional descriptors into compact binary strings, have become very popular because of their high efficiency in search and storage capacity. In the first part, we propose a multimodal retrieval method based on latent feature models. The procedure consists of a nonparametric Bayesian framework for learning underlying semantically meaningful abstract features in a multimodal dataset, a probabilistic retrieval model that allows cross-modal queries and an extension model for relevance feedback. In the second part, we focus on supervised hashing with kernels. We describe a flexible hashing procedure that treats binary codes and pairwise semantic similarity as latent and observed variables, respectively, in a probabilistic model based on Gaussian processes for binary classification. We present a scalable inference algorithm with the sparse pseudo-input Gaussian process (SPGP) model and distributed computing. In the last part, we define an incremental hashing strategy for dynamic databases where new images are added to the databases frequently. The method is based on a two-stage classification framework using binary and multi-class SVMs. The proposed method also enforces balance in binary codes by an imbalance penalty to obtain higher quality binary codes. We learn hash functions by an efficient algorithm where the NP-hard problem of finding optimal binary codes is solved via cyclic coordinate descent and SVMs are trained in a parallelized incremental manner. For modifications like adding images from an unseen class, we propose an incremental procedure for effective and efficient updates to the previous hash functions. Experiments on three large-scale image datasets demonstrate that the incremental strategy is capable of efficiently updating hash functions to the same retrieval performance as hashing from scratch.

Childhood Attention Problems and the Development of Comorbid Symptoms at the Transition to High School: The Mediating Role of Parent and Peer Relationships

Children with Attention-Deficit/Hyperactivity Disorder (ADHD) are at increased risk for the development of depression and delinquent behavior. Children and adolescents with ADHD also experience difficulty creating/maintaining high quality friendships and parent-child relationships, and these difficulties may contribute to the development of co-morbid internalizing and externalizing symptoms in adolescence. However, there is limited research examining whether high quality friendships and parent-child relationships mediate the relation between ADHD and the emergence of these co-morbid symptoms at the transition to high school. This study examines the mediating role of relationship quality in the association between ADHD and depressive symptoms/delinquent behaviors at this developmentally significant transition point. Results revealed significant indirect effects of grade 6 attention problems on grade 9 depressive symptoms through friendship quality and quality of the mother-child relationship in grade 8. Interventions targeting parent and peer relationships may be valuable for youth with ADHD to promote successful transitions to high school.

CREATING ORGANIZATIONAL CITIZENS: A QUASI-EXPERIMENTAL AND QUALITATIVE EXAMINATION OF SUPERVSIOR- AND PEER-BASED INTERVENTIONS

Research suggests that supervisors and peers can help employees make sense of what is important or expected from them at work and, thereby, shape their behaviors. In this dissertation, I examine how employees’ organizational citizenship behaviors (OCB), such as helping and voice, are differentially affected by these two sources of influence over time. In particular, I compare the relative and joint effectiveness of two field interventions to enhance OCB: (a) a role clarification intervention in which supervisors are trained to set expectations for OCB for their employees and encourage them to engage in OCB and (b) a norm establishment intervention in which peers are trained to set expectations for each other and encourage each other to perform OCB. I utilize a mixed methods approach involving a quasi-field experiment to test for changes in OCB and qualitative data to explore the theoretical mechanisms over the course of three months in a large food processing plant. I find that role clarification interventions alone have immediate positive effects on OCB, whereas norm establishment interventions alone take a longer period of time to increase OCB. In addition, in the condition where both interventions were combined, norm establishment interventions weaken the effects of role clarification earlier on; however, at later stages in time, this pattern reverses as norm establishment enhances the effects of role clarification on OCB. Through these findings, I highlight how (a) organizations seeking quick increases in citizenship might be better off focusing on supervisors as sources of influence; (b) organizations need to persist with peer-focused interventions to see positive gains; and (c) despite initial hurdles with peer-focused interventions, over time, they can lead to the highest increases in OCB when combined with supervisor-focused interventions.