Big Graph Processing : Partitioning and Aggregated Querying

Abstract : With the advent of the "big data", many repercussions have taken place in all fields of information technology, advocating innovative solutions with the best compromise between cost and accuracy. In graph theory, where graphs provide a powerful modeling support for formalizing problems ranging from the simplest to the most complex, the search for NP-complete or NP-difficult problems is rather directed towards approximate solutions, thus Forward approximation algorithms and heuristics while exact solutions become extremely expensive and impossible to use. In this thesis we discuss two main problems: first, the problem of partitioning graphs is approached from a perspective big data, where massive graphs are partitioned in streaming. We study and propose several models of streaming partitioning and we evaluate their performances both theoretically and empirically. In a second step, we are interested in querying distributed / partitioned graphs. In this context, we study the problem of aggregative search in graphs, which aims to answer queries that interrogate several fragments of graphs and which is responsible for reconstructing the final response such that a Matching approached with the initial query
Document type :
Theses
Complete list of metadatas

Cited literature [133 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-01707153
Contributor : Abes Star <>
Submitted on : Monday, February 12, 2018 - 3:35:06 PM
Last modification on : Friday, May 17, 2019 - 10:32:14 AM
Long-term archiving on : Monday, May 7, 2018 - 10:52:01 PM

File

TH2017ECHBARTHIGHIZLANE.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-01707153, version 1

Citation

Ghizlane Echbarthi. Big Graph Processing : Partitioning and Aggregated Querying. Databases [cs.DB]. Université de Lyon, 2017. English. ⟨NNT : 2017LYSE1225⟩. ⟨tel-01707153⟩

Share

Metrics

Record views

395

Files downloads

452