Profiling by Ethersource

Language is social, but language use is personal. Finding the specifics of someone's language use well enough to distinguish it from others' involves the careful modelling of linguistic behaviour to the extent that even most human readers will find this challenging in all but the most obvious cases: identifying writing style is a task which requires a fine-grained analysis of usage patterns. We use Ethersource as a basis for developing a profile for an author. Here are examples of how we can display author variation with respect to a number of classic dimensions of register variation.

These are techniques in the making which can be used to profile users to find a semantic fingerprint to accompany the stylistic fingerprint profiling mechanisms work with today.