Comments on: Sentiment Analysis of Tweets – Predicting Sentiments using Machine Learning Algorithms https://www.opencodez.com/machine-learning/sentiment-analysis-tweets-predicting-sentiments-using-machine-learning-algorithms.htm develope | share | reuse Thu, 04 Jun 2020 08:32:03 +0000 hourly 1 By: Rounak https://www.opencodez.com/machine-learning/sentiment-analysis-tweets-predicting-sentiments-using-machine-learning-algorithms.htm/comment-page-1#comment-95492 Tue, 25 Feb 2020 09:45:54 +0000 https://www.opencodez.com/?p=4522#comment-95492 Hi Naganna, sure. I think you are asking how can we label the tweets as positive, negative and neutral. I have pointed that out either we can manually label them which will be very time-consuming. Alternatively, we have used a function from ‘sentiment’ package which labels the tweets into positive, negative and neutral. This is not very accurate but for generating a basic understanding of text analytics we have worked on that in the blog.

Please let us know if you need further clarification.

]]> By: Rounak https://www.opencodez.com/machine-learning/sentiment-analysis-tweets-predicting-sentiments-using-machine-learning-algorithms.htm/comment-page-1#comment-95490 Tue, 25 Feb 2020 09:38:27 +0000 https://www.opencodez.com/?p=4522#comment-95490 Sure, Brian. Yes. This package was archived long back in December 2012. To install sentiment package will also require packages tm and Rstem(this package is also archived). You can follow the below steps to install it.
1) require(devtools)
2) packageurl <- "http://www.omegahat.net/Rstem/Rstem_0.4-1.tar.gz&quot;
3) install.packages(packageurl, repos=NULL)
4) install.packages("tm")
5) packageurl <- "https://cran.r-project.org/src/contrib/Archive/sentiment/sentiment_0.1.tar.gz&quot;
install.packages(packageurl, repos=NULL)

Please let me know if further help is needed.

]]> By: Brian https://www.opencodez.com/machine-learning/sentiment-analysis-tweets-predicting-sentiments-using-machine-learning-algorithms.htm/comment-page-1#comment-95268 Thu, 13 Feb 2020 13:19:13 +0000 https://www.opencodez.com/?p=4522#comment-95268 Hello! Thanks for a great guide. Iam trying to run the script: Flood_text_pol = classify_polarity(Flood_text, algorithm=”bayes”). It keeps returung at error:Error in classify_polarity(Flood_text, algorithm = “bayes”, prior = 1) :
could not find function “classify_polarity”. When I install package sentiment, I get thiserror: Warning in install.packages :
package ‘sentiment’ is not available (for R version 3.6.2)

Can you please help?

]]> By: Naganna Chetty https://www.opencodez.com/machine-learning/sentiment-analysis-tweets-predicting-sentiments-using-machine-learning-algorithms.htm/comment-page-1#comment-94299 Fri, 29 Nov 2019 10:43:44 +0000 https://www.opencodez.com/?p=4522#comment-94299 Dear Rounak, can you tell me how label the tweets for training purpose?

]]> By: Rounak https://www.opencodez.com/machine-learning/sentiment-analysis-tweets-predicting-sentiments-using-machine-learning-algorithms.htm/comment-page-1#comment-94083 Fri, 15 Nov 2019 07:18:51 +0000 https://www.opencodez.com/?p=4522#comment-94083 Hi DonK. Thanks for sharing your issue. Can you provide the command and share more info about the result? There are supposed to be 88 columns of data for each tweet when we use search_tweets command. Should not be any problem due to dev account setup.

]]>