LarKC’s Emanuele Della Valle to speak at NTT DoCoMo European Labs
Emanuele Della Valle will give a speach at NTT DoCoMo European Labs.
Abstract: “Building Mobile Data Mashups Using LarKC”
Is public transportation where I am? Is the event where I am the one that attract more people right now? Where are all my friends meeting? Is the traffic moving where Im going? Although the data is often available, theres no software system capable of automatically computing the answers – indeed, no system enables Mobile users even to issue such queries.
Data Mashups are a promising attempt to combine data from one or more sources into a single integrated tool.The term Mashup normally implies easy and fast integration, but it is not always true (i.e., not all the Mashups simply display events on a map). Developing a generic data Mashup engine implies a system able to cope with
- (a) representational, reasoning, and defaults heterogeneity,
- (b) data scale and distribution,
- (c) time-dependency (e.g., static vs. frequently changing info),
- and (d) noisy, uncertain and inconsistent data.
LarKC is a EU FP 7 Large-Scale Integrating Project LarKC that aims at developing a platform for massive distributed incomplete reasoning for the Semantic Web. The Urban Computing use case of LarKC is exploiting LarKC to build Mobile Data Mashups. In particular, the current focus is in answering to “Which Milano monuments can I quickly visit from here?” by developing a Mobile Data Mashup. To do so, our LarKC application combines route planning techniques (over Milano street topography) with reasoning on symbolic knowledge (e.g., linking the positions of Milano monuments taken from DBpedia to Milano street topography), traffic predictions produced by recurrent neural networks and residual street capacities produced by continuously querying Milano traffic data streams.
This talk presents the challenges the Urban Computing use case of LarKC is posing to the project and it illustrates how a Mobile Data Mashup can be develop using LarKC. An outlook of future works, with a particular emphasis on managing streaming data within the LarKC framework, is also illustrated.