Frank Rinaldo

Modeling the Dynamics of Social Networks

Modeling human dynamics responsible for the formation and evolution of the so-called social networks – structures comprised of individuals or organizations and indicating connectivities existing in a community – is a topic recently attracting a significant research interest. It has been claimed that these dynamics are scale-free in many practically important cases, such as impersonal and personal communication, auctioning in a market, accessing sites on the WWW, etc., and that human response times thus conform to the power law. While a certain amount of progress has recently been achieved in predicting the general response rate of a human population, existing formal theories of human behavior can hardly be found satisfactory to accommodate and comprehensively explain the scaling observed in social networks. In the presented study, a novel system-theoretic modeling approach is proposed and successfully applied to determine important characteristics of a communication network and to analyze consumer behavior on the WWW.

Personalized Recommendation in Interactive Visual Analysis of Stacked Graphs

We present a system which combines interactive visual analysis and recommender systems to support insight generation for the user. Our approach combines a stacked graph visualization with a content-based recommender algorithm, where promising views can be revealed to the user for further investigation. By exploiting both the current user navigational data and view properties, the system allows the user to focus on visual space in which she or he is interested. After testing with more than 30 users, we analyze the results and show that accurate user profiles can be generated based on user behavior and view property data.

Inference of Viewed Exhibits in a Metaverse Museum

This paper presents how to infer viewed exhibits in a metaverse museum from a visitors movement log. This task consists of movement-state detection and viewed-exhibit inference. For the former task, we focus on visitors fast and slow movement and discuss three differences between our previous methods and new methods: an additional parameter, speed normalization, and key-input accumulation. For the latter task, our new method focuses on the distance and angle between the visitor and each involving exhibit. According to a conducted experiment, the proposed methods could improve the performance (F-measure) by 10.3% and 4.6% for movement state detection and viewed-exhibit inference, respectively.

Evolutionary Computation for Label Layout on Unused Space of Stacked Graphs

Placing numerous objects and their corresponding labels in the stacked graph visualization is a challenging problem. In the stacked graph, different combinations of initial parameters and filtering effects yield views with hidden information, illegible labels, and unused space. The result is a tool that does not take advantage on the unused space to reveal information to the user for further investigation. We present an automatic method for label layout on the unused space in a stacked graph. An evolutionary computation (EC) is used to optimize the best label position according to legibility requirements, as well as requirements for keeping semantic relationships between labels and their representative visual objects. A number of EC experiments, as well as a usability study on label legibility, show that our proposed solution looks promising, as compared to the traditional solutions.

A Hidden Variable Approach to Analyze “Hidden” Dynamics of Social Networks

This paper deals with the statistical analysis of social networks, and it consists of two parts. First, a survey of the existing, power-law -inspired approaches to the modeling of degree distributions of social networks is conducted. It is argued, with the support of a simple experiment, that these approaches can hardly accommodate and comprehensively explain the range of phenomena observed in empirical social networks. Second, an alternative modeling framework is presented. The observed, macro-level behavior of social networks is described in terms of the individual, “hidden” dynamics, and the necessary equations are given. It is demonstrated, via experiments, that a Laplace-Stieltjes hypertransform of the distribution function of human decision-making or reaction time often provides for an adequate model in statistical analysis of social systems. The study results are briefly discussed, and conclusions are drawn.