Search

International audience; The size of available graphs has drastically increased in recent years. The real-time visualization of graphs with millions of edges is a challenge but is necessary to grasp information hidden in huge datasets. This article presents an end-to-end technique to visualize huge graphs using an established Big Data ecosystem and a lightweight client running in a Web browser. For that purpose, levels of abstraction and graph tiles are generated by a batch layer and the interactive visualization is provided using a serving layer and client-side real-time computation of edge bundling and graph splatting. A major challenge is to create techniques that work without moving data to an ad hoc system and that take advantage of the horizontal scalability of these infrastructures. We introduce two novel scalable algorithms that enable to generate a canopy clustering and to aggregate graph edges. These two algorithms are both used to produce levels of abstraction and graph tiles. We prove that our technique guarantee a quality of visualization by controlling both the necessary bandwidth required for data transfer and the quality of the produced visualization. Furthermore, we demonstrate the usability of our technique by providing a complete prototype. We present benchmarks on graphs with millions of elements and we compare our results to those obtained by state of the art techniques. Our results show that new Big Data technologies can be incorporated into visualization pipeline to push out the size limits of graphs one can visually analyze.

Recent reports have demonstrated the multi-analyte detection capability of paper networks with multiple outlets per inlet. In this report, we focus on the capabilities of 2D paper networks with multiple inlets per outlet and demonstrate the controlled transport of reagents within paper devices. Specifically, we demonstrate methods of controlling fluid transport using the geometry of the network and dissolvable barriers. Finally, we discuss the implications for higher sensitivity detection using this type of 2D paper network.

The paper studies a distributed implementation method for the BIP (Behavior, Interaction, Priority) component framework for modeling heterogeneous systems. BIP offers two powerful mechanisms for describing composition of components by combining interactions and priorities. A system model is layered. The lowest layer contains atomic components; the second layer, describes possible interactions between atomic components; the third layer includes priorities between the interactions. The current implementation of BIP is based on global state operational semantics. An Engine directly interprets the operational semantics rules and computes the possible interactions between atomic components from global states. The implementation method is a translation from BIP models into distributed models involving two steps. The first translates BIP models into partial state models where are known only the states of the components which are ready to communicate. The second implements interactions in the partial state model by using message passing primitives. The main results of the paper are conditions for which the three models are observationally equivalent. We show that in general, the translation from global state to partial state models does not preserve observational equivalence. Preservation can be achieved by strengthening the premises of the operational semantics rules by an oracle. This is a predicate depending on the priorities of the BIP model. We show that there are many possible choices for oracles. Maximal parallelism is achieved for dynamic oracles allowing interaction as soon as possible. Nonetheless, these oracles may entail considerable computational overhead. We study performance trade-offs for different types of oracles. Finally, we provide experimental results illustrating the application of the theory on a prototype implementation. © 2008 Springer-Verlag Berlin Heidelberg.