Forbes: How Machine Learning Truly Applies To Digital Identity: Defining The Problem

With the onset of digital advertising, one of the early promises was to reach specific individuals or small groups of individuals, as opposed to a large audience of a broadcast TV or radio show. To fulfill this promise, advertising and marketing technology companies, publishers and brands have been on a quest to solve digital identity.

Over the years, various identifiers or combinations of identifiers have been used as the digital identifiers of choice. These include cookies and supercookies, email addresses, UDID (unique device identifier), IMEI (international mobile equipment identity), PII (personally identifiable information such as email or phone number) advertising IDs and social media IDs (Twitter handles, Facebook IDs, etc.). For each identifier, there are pros and cons, particularly around consumer privacy, siloed data collection and the ability to market to these identifiers.

Read more here.