Information Flows in a Gallery-Work-Entertainment Space: The Effect of a Digital Bulletin Board on Social Encounters

Human Organization, July 2009, Volume 68, Issue 2, p.206-217, 2009
Information Flows in a Gallery-Work-Entertainment Space: The Effect of a Digital Bulletin Board on Social Encounters
Elizabeth Churchill, Les Nelson, Elizabeth Churchill, Les Nelson

Digital media displays are increasingly common in public spaces. Typically, these are minimally interactive and predominantly function as signage or advertisements. However, in our work we have been exploring how digital media public displays can be designed to facilitate community content sharing in civic buildings, in organizations, and at social gatherings like conferences.

While most of our installations have been within fairly formal, professional settings, in this paper we address the impact of a digital community display on interactions between the inhabitants of a neighborhood art gallery and café.

We describe the location, the display itself, and the underlying content distribution and publication infrastructure. Findings from qualitative and quantitative analyses before and after the installation demonstrate that patrons easily adopted use of the display, which was used frequently to find out more about café/gallery events and for playful exchanges.

However, despite the enthusiasm of patrons and café staff, the café owners were wary of maintaining or extending the technology. We speculate on this reticence in terms of potential for services and technologies in public space technology design.

Another publication from the same category: Machine Learning and Data Science

IEEE Computing Conference 2018, London, UK

Regularization of the Kernel Matrix via Covariance Matrix Shrinkage Estimation

The kernel trick concept, formulated as an inner product in a feature space, facilitates powerful extensions to many well-known algorithms. While the kernel matrix involves inner products in the feature space, the sample covariance matrix of the data requires outer products. Therefore, their spectral properties are tightly connected. This allows us to examine the kernel matrix through the sample covariance matrix in the feature space and vice versa. The use of kernels often involves a large number of features, compared to the number of observations. In this scenario, the sample covariance matrix is not well-conditioned nor is it necessarily invertible, mandating a solution to the problem of estimating high-dimensional covariance matrices under small sample size conditions. We tackle this problem through the use of a shrinkage estimator that offers a compromise between the sample covariance matrix and a well-conditioned matrix (also known as the "target") with the aim of minimizing the mean-squared error (MSE). We propose a distribution-free kernel matrix regularization approach that is tuned directly from the kernel matrix, avoiding the need to address the feature space explicitly. Numerical simulations demonstrate that the proposed regularization is effective in classification tasks.