What these websites do is load the website on the server and parse the received html to extract information such as the title. The probably use cURL in combination with some PREG regex.