I think you need the API to get the information in a format that is easily read by code. Scraping posts from the website interface is much more difficult and resource intensive. Also if twatter detects an IP address that is making too many requests they will probably block it.