Ranker poster (2)

I would add an slide regarding the ranker, explaining which factors TweetRank measures. Something like that:

Stochastic matrix built as a weighted sum of:

• Randomly accessing tweet j from tweet i.
• Accessing tweet j which is retweeted or replied by tweet i.
• Accessing author of tweet j mentioned by tweet i, and then accessing tweet j.
• Accessing author of tweet j followed by author of tweet i, and then accessing tweet j.
• Accessing a hashtag shared by tweet i and tweet j, and then accessing tweet j.

This matrix represents the total probability of accessing tweet j from tweet i.

TweetRank is the eigenvector of this stochastic matrix.