A data model helps illustrate the way groups of information within a data set relate to one another. Data models can represent all sorts of information, from customer call-center databases to apartment-rental data used by real estate agents. Multiple models can be applied to individual data sets in order to view the information in a variety of ways and ensure all relationships are represented.

Sam Wilcke, director of analytics at Marin Software, on data models

Continue reading at AdAge.com

Real-time modeling is an approach to audience modeling that employs information that can be put to use immediately after being collected.

Keith Gooberman, VP-trading and platform operations, at Varick Media, on Real-Time Modeling:

Real-time modeling would be trying to identify an audience in real-time so you can quickly act upon it. For example, noticing a jump in activity driven to your website in the morning (due to some specific, searchable term) might influence you to contextually target the word using display advertising immediately.

Continue reading at AdAge.com

Quantitative data is data that can be measured numerically. Things that can be measured precisely -- rather than through interpretation -- such as the number of attendees at an event, the temperature in a given location, or a person's height in inches can be considered quantitative data. Its foil -- qualitative data -- requires a subjective decision in order to be categorized or measured.

Chris Deiner, SVP analytics at AbsolutData, on Quantitative Data:

Continue reading at AdAge.com