Data Mining Twitter to Improve Automated Vehicle Safety (04-098)

Project Description: Research question: Investigating the effects of Automated Vehicle (AV) events on social media user behavior. Data collection: The dataset was collected in February, 2020 from Twitter’s website. Data scope: The dataset consists of tweets published within a twenty-day window centering on 12 AV-related events from 2014 to 2019, including crashes and technology advancements (see Table 2). These events were selected by the research team from news headlines. The tweets were identified using a keyword search method (see Table 1) around the twenty-day window. The tweets were collected from Twitter’s website through web-scraping. Users who have approved access via Twitter may use the tweet ID information in the dataset to retrieve the tweet content and meta-data. Please see Data Description file for tables. Data Specification: Please see Data Description file for more information.