There are pros and cons of using self-declared data from social media users and trying to understand where fraud might be happening. There are privacy concerns to be dealt with. In my discussions of small sample sizes, it would appear that more people are  amiable to data being used to catch fraud rather than marketing or collection and such other purposes. On the other hand unstructured data from social media can often help understand the true “identity” – which is the best indicator of potential fraudulent behavior in a non-face-to-face transaction. Danielle Kucera has pointed here to some interesting possibilities….

Bloomberg article