|
Name |
Olivera Novović |
|
Affiliation |
Institute BioSense, Novi Sad, Serbia |
|
Corresponding WG |
WG 1, 4 |
|
Grant research topic |
Big Data frameworks for processing mobile phone data |
|
Hosting Institution |
Aristotle University, Thessaloniki, Greece |
|
Period |
2017-02-19 to 2017-03-05 |
|
Summary of the scientific report |
At the beginning of the visiting period, Prof. Apostolos N. Papadopouls made a short introductory lecture about Apache Spark and programing language Scala. We also discussed specifics of Spark’s GraphX library for graph mining. The following days we discussed possible algorithms that could be applied on mobile phone data graphs. We also spent some time setting up the environment IntelliJ IDEA for programing in Scala. We generated top10 page rank results using Scala and Spark on the graphs that were previously prepared. During the discussion about algorithms that could be applied we decided based on the nature of mobile phone data, to try community detection. With the exceptional assistance of Prof. Papadopouls I went over theory basics on modularity based approach for community detection algorithms. In the next few days I succeeded to generate communities for the mobile phone data graphs using Louvain algorithm. |
|
Personal notes |
It was a great pleasure working in DeLab together with Prof. Papadopouls. I’ve been provided with an office at the DeLab (Data Engineering Laboratory), Department of Informatics, Aristotle University of Thessaloniki. It was a really great experience and the knowledge that I gained during my visit is of exceptional value for my future research. |
