Software |01 Ride-sharing matcher RSM
RSM enables any rideshare participant to query a database of advertised trip schedules giving as input his intended journey. At query time, the engine first infers a time window, a path (for prospective drivers), and reachable perimeter (for potential passengers), before returning from the advertised trip schedules, the feasible ride-matches that are both geographically and temporally consistent with the querying user journey.
Software |01 RSM
Software |02 SaKey
Software |02 SaKey
SaKey is a scalable method for the discovery of almost-keys in large RDF data sets. A set of properties is a key if it uniquely identifies any object in a RDF data set. A set of properties is a k-almost-key if it allows at most k exceptions preventing the set of properties to be a key. K-almost-keys are particularly helpful for discovering same as link in the context of geographically distributed and possibly unclean datasets. This software has been developed with Danai Symeonidou.
03 M2DT &
Software |03 M2DT & M3DJ
M2DT - (Minimal Diagnoses by Distributed Tree) is a decentralized incremental algorithm for diagnosing distributed systems (available)
M3DJ - (Minimal Distributed Diagnoses by Distributed Join-tree) is an improvement of M2DT that is able to diagnose peer-to-peer systems respecting privacy
Software |04 T.E.
Software |01 Token Elimination T.E.
TE is a distributed tree decomposition algorithm that organizes and improves the distributed reasoning in a network (available). Our experiments on different types of small world graphs (which are known to model the interactions of real world applications such as WWW or ISCAS Circuits) showed that TE produces Tree Decomposition of better quality decompositions (i.e. smaller widths than classical methods in DCSP or DCOP and is competitive to the centralized method Bucket Elimination.
Software |02 AGSWRI
AGSWRI- (Automatic Generator of Small World Instances) is an automatic generator of graphs and systems descriptions having a small world structure. This generator creates graphs instances following Watts and Strogatz (WS) model but also Barabasi and Albert (BA) model. BA graphs are known to be heterogeneous and represent the interaction of the real world applications such that WWW while WS graphs are more homogeneous and represents the interaction of ISCAS circuits. Using the small world graph structure, we add a local CNF description on the top of each node.
Software |05 AGSWRI
Just a sample of my work. To see more or discuss possible work >>