Sarah Michele Rajtmajer

The Rich Get Richer: Brain Injury Elicits Hyperconnectivity in Core Subnetworks

There remains much unknown about how large-scale neural networks accommodate neurological disruption, such as moderate and severe traumatic brain injury (TBI). A primary goal in this study was to examine the alterations in network topology occurring during the first year of recovery following TBI. To do so we examined 21 individuals with moderate and severe TBI at 3 and 6 months after resolution of posttraumatic amnesia and 15 age- and education-matched healthy adults using functional MRI and graph theoretical analyses. There were two central hypotheses in this study: 1) physical disruption results in increased functional connectivity, or hyperconnectivity, and 2) hyperconnectivity occurs in regions typically observed to be the most highly connected cortical hubs, or the “rich club”. The current findings generally support the hyperconnectivity hypothesis showing that during the first year of recovery after TBI, neural networks show increased connectivity, and this change is disproportionately represented in brain regions belonging to the brains core subnetworks. The selective increases in connectivity observed here are consistent with the preferential attachment model underlying scale-free network development. This study is the largest of its kind and provides the unique opportunity to examine how neural systems adapt to significant neurological disruption during the first year after injury.

Identification and characterization of cyberbullying dynamics in an online social network

Cyberbullying is an increasingly prevalent phenomenon impacting young adults. In this paper, we present a study on both detecting cyberbullies in online social networks and identifying the pairwise interactions between users through which the influence of bullies seems to spread. In particular, we investigate the role of user demographics and social network features in predicting how users will respond to a cyberbullying comment. We characterize the influencer/influenced relationship by which a user who has no history of abuse observes a peer engaging in bullying and follows suit. To our knowledge, this is the first effort modeling peer pressure and social dynamics with analytical models. We validate our models on two distinct social network datasets, totalling over 16, 000 posts. Our results offer insight into the dynamics of bullying and confirm social theories on the power of peer groups in the cyberworld. A full version of this paper is available on arXiv.org.

A voxelwise approach to determine consensus regions-of-interest for the study of brain network plasticity

Despite exciting advances in the functional imaging of the brain, it remains a challenge to define regions of interest (ROIs) that do not require investigator supervision and permit examination of change in networks over time (or plasticity). Plasticity is most readily examined by maintaining ROIs constant via seed-based and anatomical-atlas based techniques, but these approaches are not data-driven, requiring definition based on prior experience (e.g., choice of seed-region, anatomical landmarks). These approaches are limiting especially when functional connectivity may evolve over time in areas that are finer than known anatomical landmarks or in areas outside predetermined seeded regions. An ideal method would permit investigators to study network plasticity due to learning, maturation effects, or clinical recovery via multiple time point data that can be compared to one another in the same ROI while also preserving the voxel-level data in those ROIs at each time point. Data-driven approaches (e.g., whole-brain voxelwise approaches) ameliorate concerns regarding investigator bias, but the fundamental problem of comparing the results between distinct data sets remains. In this paper we propose an approach, aggregate-initialized label propagation (AILP), which allows for data at separate time points to be compared for examining developmental processes resulting in network change (plasticity). To do so, we use a whole-brain modularity approach to parcellate the brain into anatomically constrained functional modules at separate time points and then apply the AILP algorithm to form a consensus set of ROIs for examining change over time. To demonstrate its utility, we make use of a known dataset of individuals with traumatic brain injury sampled at two time points during the first year of recovery and show how the AILP procedure can be applied to select regions of interest to be used in a graph theoretical analysis of plasticity.

A hybrid epidemic model for deindividuation and antinormative behavior in online social networks

With the increasing popularity of user-contributed sites, the phenomenon of “social pollution”, the presence of abusive posts has become increasingly prevalent. In this paper, we describe a novel approach to investigate negative behavior dynamics in online social networks as epidemic phenomena. We show that using hybrid automata, it is possible to explain the contagion of antinormative behavior in certain online commentaries. We present two variations of a finite-state machine model for time-varying epidemic dynamics, namely triggered state transition and iterative local regression, which differ with respect to accuracy and complexity.We validate the model with experiments over a dataset of 400,000 comments on 800 YouTube videos, classified by genre, and indicate how different epidemic patterns of behavior can be tied to specific interaction patterns among users.

Non-negative sparse autoencoder neural networks for the detection of overlapping, hierarchical communities in networked datasets

We propose the first use of a non-negative sparse autoencoder (NNSAE) neural network for community structure detection in complex networks. The NNSAE learns a compressed representation of a set of fixed-length, weighted random walks over the network, and communities are detected as subsets of network nodes corresponding to non-negligible elements of the basis vectors of this compression. The NNSAE model is efficient and online. When utilized for community structure detection, it is able to uncover potentially overlapping and hierarchical community structure in large networks.

An Ultimatum Game Model for the Evolution of Privacy in Jointly Managed Content

Content sharing in social networks is now one of the most common activities of internet users. In sharing content, users often have to make access control or privacy decisions that impact other stakeholders or co-owners. These decisions involve negotiation, either implicitly or explicitly. Over time, as users engage in these interactions, their own privacy attitudes evolve, influenced by and consequently influencing their peers. In this paper, we present a variation of the one-shot Ultimatum Game, wherein we model individual users interacting with their peers to make privacy decisions about shared content. We analyze the effects of sharing dynamics on individuals’ privacy preferences over repeated interactions of the game. We theoretically demonstrate conditions under which users’ access decisions eventually converge, and characterize this limit as a function of inherent individual preferences at the start of the game and willingness to concede these preferences over time. We provide simulations highlighting specific insights on global and local influence, short-term interactions and the effects of homophily on consensus.

Dynamic data-driven sensor network adaptation for border control

Given a specific scenario for the border control problem, we propose a dynamic data-driven adaptation of the associated sensor network via embedded software agents which make sensor network control, adaptation and collaboration decisions based on the contextual information value of competing data provided by different multi-modal sensors. We further propose the use of influence diagrams to guide data-driven decision making in selecting the appropriate action or course of actions which maximize a given utility function by designing a sensor embedded software agent that uses an influence diagram to make decisions about whether to engage or not engage higher level sensors for accurately detecting human presence in the region. The overarching goal of the sensor system is to increase the probability of target detection and classification and reduce the rate of false alarms. The proposed decision support software agent is validated experimentally on a laboratory testbed for multiple border control scenarios.

Positive and negative behavioral analysis in social networks.

Use of online social networks has grown dramatically since the first Web 2.0 technologies were deployed in the early 2000s. Our ability to capture user data, in particular behavioral data has grown in concert with increased use of these social systems. In this study, we survey methods for modeling and analyzing online user behavior. We focus on negative behaviors (social spamming and cyberbullying) and mitigation techniques for these behaviors. We also provide information on the interplay between privacy and deception in social networks and conclude by looking at trending and cascading models in social media. WIREs Data Mining Knowl Discov 2017, 7:e1203. doi: 10.1002/widm.1203

Invited Paper: A Model of Paradoxical Privacy Behavior in Online Users

Recent privacy research has identified some paradoxical behavior that online users often display, such as claiming to be conservative and then actually oversharing. In particular, recent work by Brandimarte et al. discussed an apparent privacy paradox in online interactions. According to the study, users tend to over-disclose private information if they perceive to have control over the disclosure of their content. In this paper, we formally explain this apparent paradox in user privacy behavior as a straightforward optimization of comfort with sharing and perceived control. We describe the interests of a social network site in managing user privacy options. Namely, a site seeks to maximize perceived user control, while managing costs associated with providing that control. Furthermore, we extend the discussion for the case of ynamic time, and study an optimal control problem for the site as it tries to allocate resources toward user privacy control.

A Hybrid Epidemic Model for Antinormative Behavior in Online Social Networks

In this paper, we describe a novel approach to investigate negative behavior dynamics in online social networks as epidemic phenomena. We present a finite-state machine model for time-varying epidemic dynamics, and validate this model with experiments over a large dataset of Youtube commentaries, indicating how different epidemic patterns of behavior can be tied to specific interaction patterns among users. A full version of this paper is available on arXiv.org.