In this paper, we describe our experiments using Latent Dirichlet Allocation (LDA) to model images containing both perceptual features and words. To build a large-scale image tagging system, we distribute the computation of LDA parameters using MapReduce. Empirical study shows that our scalable LDA supports image annotation both effectively and efficiently.
In bulk heterojunction photovoltaic systems both the open-circuit voltage as well as the short-circuit current, and hence the power conversion efficiency, are dependent on the optical bandgap of the electron-donor material. While first-principles methods are computationally intensive, simpler model Hamiltonian approaches typically suffer from one or more… (More)
In recent years, the rapid development of industrialization and urbanization has caused serious environmental pollution, especially particulate pollution. As the "Earth's kidneys," wetland plays a significant role in improving the environmental quality and adjusting the climate. To study how wetlands work in this aspect, from the early autumn of 2014 to… (More)