2

I am doing text mining on twitter data. The problem I am facing is I have data in 27 languages, but I want all the tweets to be converted to English. How can I do that using tm package or translate package?

RHA
  • 3,677
  • 4
  • 25
  • 48
Kavya
  • 31
  • 1
  • 6
  • 2
    If you got the input languages, you can use the Bing API with `translateR`: http://stackoverflow.com/questions/36598241/translating-parts-of-datasets-with-r/36601465#36601465 – lukeA Apr 14 '16 at 10:53
  • why are you trying to translate tweets into English? I doubt you will have much luck with a translation engine or API as tweets are a very informal linguistic register, characterized by slang, abbreviations, acronyms, emoticons, etc. and most translators are likely trained on more formal or canonical registers/varieties of language. – Brandon Loudermilk Apr 14 '16 at 11:30
  • Why not cut the linguistic task down by first processing in native languages and then turning to the 'what does this mean' in English part of the problem. – Chris Apr 14 '16 at 14:41
  • @lukeA Thanks alott. The link you have mentioned was very helpful, I was trying it from last 3 days, finally got some desired output without any error...But still I have left with one problem - the translate command change the text from one language to another what if I want to change different languages to only one language simultaneously. One way is I have to run this function many times by keep on changing the source language but it will be little time consuming, is there is any other way to do this..Thanks once again..!! – Kavya Apr 19 '16 at 07:20

0 Answers0