A Storm-Based Tag Cloud Platform for Multiple SNS Users
Siwoon Son, Dasol Kim, Myeong‐Seon Gil, Yang‐Sae Moon · 2017
In general, social network service (SNS) has many difficulties in collecting, storing, and analyzing the data in real time because it has the characteristic of big data generated quickly by combining structured and unstructured data. In this paper, we propose a scalable visualization technique that can be applied to real-time SNS data analysis. We first draw out the problems of the existing SNS visualization system and then define the requirements to solve those problems. For resolving these requirements, we use Apache Storm to collect and aggregate Tweet SNS data, which are generated at a high speed, and we visualize the aggregated result in a dynamic tag cloud. To do this, we design and implement a Web interface that allows multiple users to input keywords and visualize their aggregate results in real time. Also, we empirically compare the results of dynamic tag clouds to ensure that they are visualized correctly. Through this research, the users can intuitively understand how the topics of interest are changing in SNS in real time, and the visualization results can be applied to other useful services such as subject trend analysis and customer needs identification.