Twitter tweet scraping for the purposes of brand sentiment analysis
Twitter data scraped without using the Twitter API.
This code allows you to first selection a location (lat & long) and radius of search size. Selecting a keyword, such as "Toyota", will find the tweets that contain the word/hashtag in the body of the text.
In addition to the body of the tweet, I also am grabbing the user location, handle, display name, and time of the tweet.
To aid the end goal of NLP-based brand sentiment analysis, I also use the langdetect package to detect the language of the tweet. Additionally, I removed tweets from the brand's official Twitter handle in order to avoid a positive bias in terms of sentiment.