I'd like to explore Twitter's API to do sentiment analysis via the emojis used in tweets containing certain keywords or hashtags. (Like what PRISMOJI do: https://twitter.com/PRISMOJI?s=09)
But I've never done this before - any pointers on the script?
I'm using rtweet to explore the API, but open to other tools - but hopefully in R.
UPDATE
So I've written some code and solved my original issue. Woop!)
install.packages("rtweet")
library(rtweet)
install.packages("tidyverse")
library(tidyverse)
install.packages("tidytext")
library(tidytext)
insrall.packages("httpuv")
library(httpuv)
api_key <- 'api key entered here'
api_secret <- 'api secret entered here'
access_token <- 'token entered here'
access_token_secret <- 'key entered here'
token <- create_token(
app = "my app name",
consumer_key = api_key,
consumer_secret = api_secret
access_token = access_token,
access_secret = access_token_secret)
tweets <- search_tweets(q ="#CleanDishwasher", n = 18000, include_rts = FAKSE, `-filter` = "replies", lang = "en")
## get coordinates associated with the following addresses/components
write_as_csv(tweets, "tweets.csv")
sydney <- lookup_coords("sydney", "country:AU")
## pass a returned coords object to search_tweets
sydney_tweets <- search_tweets(geocode = sydney)
##find top emojis
library(emo)
sydney_tweets %>%
mutate(emoji = ji_extract_all(text)) %>%
unnest(cols = c(emoji)) %>%
count(emoji, sort = TRUE) %>%
top_n(10)
When setting my access keys and tokens, I had mistakenly used access_token_secret rather than access_secret.