Collecting Twitter data using Twitter API

Get your Twitter Developer Account

First, you have to set up a Twitter Developer Account. Go to https://developer.twitter.com/en/docs/twitter-api/getting-started/getting-access-to-the-twitter-api and sign up.

Note: Use at your own risk. Be ethical about user data.

Required libraries

library(rtweet)
library(dplyr)
library(ggplot2)
library(tidytext)
library(stringr)

Create a Twitter App and generate necessary keys and tokens

In order to create an app, you need to log in to Twitter Developer Account and click on “+Create App”.

Afterwards you need to generate the necessary keys/tokens to access the official Twitter API (click on regenerate/yellow).

I would advice to store those in your R script. Note: These are fake example values. Please replace those with your own keys.

api_key <- "yourownkey"
api_secret_key <- "xyz123456"
access_token <- "yourowntoken"
access_token_secret <- "xyz123456P"

Authenticate via web browser

Note: The name of the app is an example, exchange for your own app name.

create_token(
  app = "Example_app",
  consumer_key = api_key,
  consumer_secret = api_secret_key,
  access_token = access_token,
  access_secret = access_token_secret)

Start the madness

Here the collection process starts. We want to collect n= 500 tweets including the hashtag #starwars

tweets_starwars <- search_tweets(q="#starwars",
                               n = 500)

View the first 5 rows of the text in the dataframe

head(tweets_starwars$text, n = 5) 

Count the most frequently used words and plot them (using ggplot2)

tweets_starwars_only_text <- tweets_starwars %>%
  dplyr::select(c(5)) %>% # select the "text" column
  unnest_tokens(output = word, input = text)%>% # tokenization
  anti_join(stop_words)%>% # remove stopwords
  filter(!str_detect(word, "[:punct:]|[:digit:]")) #remove punctuation and numbers
  

tweets_starwars_only_text %>% 
  count(word, sort = TRUE) %>%
  mutate(word = reorder(word,n)) %>%
  na.omit() %>%
  top_n(15) %>% # take top 15 words
  ggplot(aes(x = word,y = n)) +
  geom_col() +
  coord_flip() +
      labs(x = "Most frequent words",
      y = "Count",
      title = "Words of star wars tweets ")

← Previous Post Next Post →

Collecting Twitter tweets

Using Twitter API to collect Starwars tweets