• We want to do a sentiment analysis on the World News subreddit comments for the different posts relating to recent news from all over the World
• It would be interesting to see how the Reddit community is reacting to recent news in light of some attention-seeking events that have happened recently.

• We will use RedditExtractoR, tidyverse, and other relevant libraries; no secondary data source.
• Data can be retrieved by supplying the search arguments like so:

reddit_links <- reddit_urls(
  subreddit = "worldnews",
  page_threshold = 10)

reddit_thread <- reddit_content(reddit_links$URL)

• Data retrieved as a data frame.
• Initial data explorations shows no violation of any tidy rules.
• Since we are dealing with text data, we will need to clean and transform the text to remove stopwords, stemming, etc.

• Data exploration for comments based on posting times and number of comments in threads
• Convert the comment texts to vectors, so we can perform a text analysis on them.
• N-grams, TF-IDF, Sentiment Analysis
• Use Sentiments Analysis
  • Positive / Negative
  • Emotion
  • Numeric

• Wordcloud to get most frequent phrases, sentiment analysis results
• Graphs, histograms for domains
• Sentiment score/count plot