Objectives of the AMANDA project, based on its organization into workparts and tasks.
This WorkPart aims at developing innovative algorithmic methodologies for handling very large data sets. We will pursue the design of novel algorithms for selected data-intensive applications in the biomedical domain and the development of tools for the analysis of the dynamic behavior of software systems dealing with massive data. As a complementary endeavor, we will explore foundational issues such as resiliency to faults, resource-performance tradeoffs, visualization of streamed graphs, and space-efficient storage and retrieval. Particular emphasis will be posed on computational models based on emerging paradigms for big data processing (e.g., map-reduce). These objectives will be pursued in the following two tasks:
WorkPart 2 aims at developing novel algorithmic strategies for dealing with large data sets having a networked and often dynamic structure. We will focus on two main issues: computing structural properties of massive networks, which is a prerequisite to conceive methodologies for quickly mining information from the data, and designing ad-hoc algorithms and visual interfaces for supporting the mining process. Namely, the research approach is schematized in the following pipeline: Massive and Evolving Networked Data -> Compute Structural Properties -> Design Algorithms and Interfaces for Network Mining. Reflecting this scheme, WP2 consists of two tasks.
|Webmaster: Maurizio Patrignani (firstname.lastname@example.org)|