Do you know a technique to find spatial indicators using the package TwitteR in R? For instance if I have to find the word "dog" using the package twitteR, but I miss the coordinates, there is the possibility to find the localization of the tweets thanks to some indications included in the time-zone, websites, UTC24offset,text message. How can I say to R to detect my list of Tweets contains the word "dog" using the spatial indicators just mentioned?
-
There was a recent post on twitter's blog that the USGS is using twitter to verify earthquakes : blog.twitter.com/2015/usgs-twitter-data-earthquake-detection ... you might try checking if the people/projects mentioned have any presentations, or try contacting them.– JoeOct 8, 2015 at 18:26
Add a comment
|
1 Answer
You could use the utc_offset or time_zone keys of the user section of a tweet.json (I trimmed the user dict) - full tweet description:
"user":{
"utc_offset":-28800,
"time_zone":"Pacific Time (US & Canada)",
"created_at":"Wed May 23 06:01:13 +0000 2007"
}
But this is based on the user account and not on a specific tweet. Still, it gives you a probabilistic geolocation for non-geotagged tweets.