dc.description.abstract |
Social media has a great influence on our social life and extant literature finds that
Twitter trends can capture the electoral sentiment. Twitter is a micro-blogging app,
is the latest gift of globalization for social networking. We have gathered around
0.3 million tweets from 11 February 2019 to 31 March 2019 from the twitter accounts
of about 200 political leaders belonging to the ruling as well as opponent parties.
The research mainly aims to focus on social media data, data on political landscape,
political inclination of different political groups, data mining of politically relevant
tweets, data extraction with the help of Java and Twitter API. Several Graphs over
different time periods have been generated based on a sense of their common contexts
in the tweets during such periods. These contexts, in turn, identify the important
events occurring during the same period. We have made two significant
observations based on the above analysis. First, we found that the degree distributions
of the generated graphs, each of whose nodes represent the political members
only, during different periods corresponding to the events, follow power-law distribution.
This result closely resembles real world complex network characteristics.
Second, based on the data, it has been observed that with respect to such influential
incidents like Pulwama explosion and killing or the Balakot airstrike when the ruling
party was using, in their tweets, certain type of words related to the sensational
events, the opposition parties’ tweet accounts show that such words were mostly
absent and other words were more prevalent, a fact which may have crucial political
significance in the context of the Indian General Election in determining which type
of views on these issues may have played determining role in the final voting. |
en_US |