Over the past decade, distributed computing has experienced a dramatic scale shift, with respect to size, geographical spread and volume of data. Meanwhile, Internet has moved into homes, creating tremendous opportunities to exploit the huge amount of resources at the edge of the network. Search engines that navigate this universe are astonishingly powerful and rely on sophisticated tools to scan and index the network. However, the network contains far more than just the pages such systems can index. There is a tremendous potential in leveraging these new kinds of information to empower individuals in ways that Internet search will never be able to offer. This reveals striking evidence that navigating the Internet goes beyond traditional search engines. Complementary and different means to navigate the digital world are now required. The objective of GOSSPLE is to provide an innovative and fully decentralized approach to navigate the digital information universe by placing users affinities and preferences at the heart of the search process. GOSSPLE will turn the network into a self-organizing federation of overlapping sub-networks, capturing on the fly the interactions and affinities observed in real life and fully leveraging the huge resource potential available on edge nodes. GOSSPLE will provide a set of fully decentralized algorithms to efficiently search, dynamically index and asynchronously disseminate information to interested users based on their preferences and (implicit) recommendations. Building up upon the peer to peer communication paradigm and harnessing the power of gossip-based algorithms, GOSSPLE will yield a disruptive way of programming distributed collaborative applications. Our goal is ambitious: impose the GOSSPLE approach as a fully decentralized, collaborative and scalable, yet complementary, alternative to traditional search engines to fully exploit the capabilities of the digital universe.
Field of science
- /natural sciences/computer and information sciences/internet
Call for proposal
See other projects for this call