ANALYSIS OF TWITTER DATA BY EXTRACTING THE TWEETS ON DEMONETIZATION
Keywords:
Tweets, Demonetization, Analytics Of Data, TwitterAbstract
This report is on how to perform analytics on tweets of particular topic. In this we are going to review the tweets on demonetization. This report also presents various findings and visualisations while analysing the tweets extracted/provided by twitter on the topic demonetization and also the context is confined to analysis of twitter data.
References
Rathee, A. Demonetization in India Twitter Data. Retrieved from Kaggle: https://www.kaggle.com/arathee2/demonetizationin-india-twitter-data.
Social Samosa. Social Samosa. Retrieved from socialsamosa.com: https://www.socialsamosa.com/2016/11/data-social-mediareactions-demonetization/Appendix A - Techniques in R that are used to perform analytics on Twitter data
To get the word cloud, use library(”wordcloud”) and to generate the word cloud use wordcloud(tweets.text.corpus,min. freq = 2, scale=c(7,0.5),colors=bre wer. Pal (8,” Dark2”), random.color= TRUE,random.order = FALSE, max.words =250).
We need to remove links,punctuation marks,blank spaces ,tabs and the user name from the obtained tweets as part of data cleaning, because these are unnecessary for analytics. 3. To remove something from a text we need to use the funtion gsub. So in case we want to clean the html links we use new txt¡-gsub(” http[ˆ[:blank:]]+”,””,ini txt).
To get the plot of hierarchical clustering on the words extracted from the tweets,clust¡-hclust(dis,method=”ward.D2”), plot(clust,cex=0.9,hang=1,main= ”Dendogram”), rect.hclust(clust,k=3).
The algorithm used to classify emotion and polarity is bayesclassifier.In R we use class emo = classify emotion(some txt, algorithm=”bayes”, prior=1.0) and for polarity class pol = classify polarity(some txt, algorithm=”bayes”).
Downloads
Published
Issue
Section
License
Copyright (c) 2016 Lakshmi Namratha Vempaty (Author)
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
