SYSTEM AND METHOD FOR DETERMINING USER IDENTITY FRAUD USING SIMILARITY SEARCHING

摘要

The invention (100) uses profile searching to search data from one data feed against another database, in order to find similarities between the two sets of data. A new user profile data feed (101) is transmitted to a similarity search engine (SSE) batch search component (103). The data feed (101) can come from many different sources where data can be acquired. The source of the data feed (101) may comprise, for example, Internet web sites, relational databases, data entry systems, and hierarchical databases. The new user profile data feed (101) is used as an anchor for search component (103) to perform a similarity search against hierarchical documents in a denied users profile database (102). The search component (103) formulates the new user profile data feed (101) into a search request that will be used against the denied user profile database (102). Once the search component (103) completes its search of the denied user profile database (102), a similarity search result set is returned. The similarity search result is examined by a match component (104) that determines whether a profile match exists. If the similarity search result set meets the specified profile match tolerance level, it is forwarded to the positive response component (105). If the similarity search result does not meet the specified tolerance level, then the new user profile is forwarded to the negative response component (106).