I began this analysis to identify whether I would get delayed on the way to work or not using Washington DC's WMATA. I began by pulling data from WMATA's API and pulling data from Twitter for weekday travel commuting. I noticed there was not a huge overlap between Tweets and WMATA's delay alerts. My hypothesis was that when there were delays, people would tweet more. However, this was not the case.
Some future work will involve:
- how to accurate create 'unhappiness' scores
- how to get additional twitter data (by adding different hashtags for example)
- create my own sentiment analysis for tweets