One thing I've been contemplating is the extent to which feedback on twitter can alter a user's opinion or focus on a given topic.
Ex.
- make a bot network of 100 bots.
- identify some accounts, split them into control and experimental groups
- choose a topic, such as "NFL"
- every time someone in experimental group tweets something about the NFL, like or retweet it a dozen times from a random slice of your bots
- do this for a few weeks
- at end of few weeks, does experimental group tweet about NFL more frequently than control? Has the sentiment within their tweets gotten any more extreme?
Oh cool, this is something I've been thinking about too. Please, if you feel you have the time, do email me (see my bio) with some lessons if you manage to build something.
Ex.
- make a bot network of 100 bots.
- identify some accounts, split them into control and experimental groups
- choose a topic, such as "NFL"
- every time someone in experimental group tweets something about the NFL, like or retweet it a dozen times from a random slice of your bots
- do this for a few weeks
- at end of few weeks, does experimental group tweet about NFL more frequently than control? Has the sentiment within their tweets gotten any more extreme?