I have one question about something ,I'm applying Topic Modeling - LDA in R language and I used it to determine the topics of the user's tweets , I notice something when I'm using this command to get the tweets ( the last 500 tweets without retweets) sometime I got 500 and sometime I got 130 and sometimes I got 45 for different users but with the same n=500 , what is the factor behind that
tweets <- userTimeline("BarackObama",n=500,includeRts = FALSE)
also if I want to remove the name of user from tweets how I can do that in R because I noticed the accuracy of clustering Topics it's not good , I don't know how we can measure the percentage of accuracy for this thing in R . Last question about if the TwitterR support arabic tweets or not !
Thanks.