Estimating gender in a mobile environment
We describe an algorithm to identify and target end-users based on their gender in a true big-data environment. We are utilizing a large number of data points originating from hundreds of millions of devices and thousands of applications daily. We use this data to derive a metric related to the probability of each individual device belonging to a female or male person. While such approaches are not new, the implementation in a real-time, big-data environment is challenging and requires continuous investment in data-science know-how and hardware set-up. Consequently, this whitepaper ends up with a claim for a neutral validation solution to justify those investments from AdTech companies and therefore higher CPM prices for the industry.
DOWNLOAD