Matching features

Cluster analysis enables identifying a given user or user group according to features, that can include age, geographic location, education level, etc. It is a data mining technique that is in use in marketing to segment a dataset and, for example, send a message to the right target for that product or service (young people, mothers, pensioners, etc.). The variable combinations are endless and make cluster analysis more or less selective according to the search requirements.


Matching similarity

Similarity is the underlying principle for making product recommendations (identifying people who are alike in terms of the products they have purchased or have liked). Online retailers such as Amazon use similarity to provide recommendations of similar products. When expressions like “People who like A also like B” appear, similarity has been applied. Similarity matching tries to recognise similar individuals based on known information about them.

Similarity matching is also an approach where attacks rely on similar features between a target dataset and auxiliary information to perform the matching. For example, mobility traces (WiFi contacts, Instant Message contacts) can be used to find distance similarities that are then matched with the help of statistical predictors. Similarities can be matched between social data and mobility traces data, and between resume and tweets, and some techniques use node similarity to match graphs in social networks.

Matching statistics

Matching statistics maps datasets statistically and relies on the unique features of users’ data (intrinsic characteristics, such as interests and political views), to perform matching operations that can lead to users being identified and tracked.