About the paper
| Title |
Architecture and Algorithms for Filtering Tweets Based on Chosen Countries and Cities |
| Book |
ICIST 2015 Proceedings |
| Pages |
402-407 |
| Abstract: In this paper we present an algorithm for
filtering Twitter data based on the tweet geographic
location. Desired geographic locations are provided as a set
of parameters, and different properties of a tweet are
considered to determine the location. A user may also
choose the number of threads and amount of memory used
in filtering process. In this way, the user may fine-tune the
algorithm performance. Filtered data are stored in the
Hadoop distributed file system which runs on a 16 nodes
cluster. Therefore, the analysis may be executed in a
distributed environment. We also present system
architecture which supports the filtering process and
analysis of filtered data. |
Full citation
Igić, N.,
Dimitrieski, V.,
Luković, I.,
Aleksić, S.,
Čeliković, M.
Architecture and Algorithms for Filtering Tweets Based on Chosen Countries and Cities. In:
Zdravković, M.,
Trajanović, M.,
Konjović, Z.
(Eds.)
ICIST 2015 Proceedings, pp.402-407, 2015
Keywords
Data and text mining
Social networks