Exploring Linkability of User Reviews

摘要: Large numbers of people all over the world read and contribute to various review sites. Many contributors are understandably concerned about privacy in general and, specifically, linkability their reviews (and accounts) across multiple In this paper, we study community-based reviewing try answer question: what extent ”anonymous” linkable, i.e., highly likely authored by same contributor? Based on a very large set from one popular site (Yelp), show that high percentage ostensibly anonymous can be accurately linked authors. This is despite fact use simple models equally features set. Our suggests reliably expose identities reviews. has important implications for cross-referencing accounts between different Also, techniques used our could adopted sites give feedback