Demographics

  • Article

You use the Demographics model configuration to build analysis models that predict the demographic characteristics that your users are likely to share. Using this configuration, you can predict missing properties a user failed to enter during the user registration processes on your site.

Following are the suggested demographic property/values.

  • Age. The age of the user. The age can be a range of numbers or one specific number.
  • Occupation. Occupation of the user. It is highly recommended that you predefine occupation groups to avoid synonyms and spelling errors. For example, even though a software engineer and a developer perform the same work, the Predictor resource has no way of knowing that unless you define and consistently use a value name for Occupation.
  • MaritalStatus. Marital status of the user. For example, single, married, divorced, and so forth.
  • Gender. Male or female.
  • Income. The income of the user. It is highly recommended that you predefine income groups.
  • Education. Educational level of the user. This property can either be a number corresponding to the amount of years a user has studied, or a level of education, such as, High School, B.A., M.S., Ph.D, and so forth.
  • Children. The number of children the user has.
  • FamilyRole. The role of the user in a family.
  • Region. Any type of regional location. This property can vary from site to site. On international sites, you might choose to designate country/regions. On sites that are not international, you might choose to use states. Moreover, local sites can use cities or counties. If you wish to use country/region, state, and city, you can create strings such as ''USA.WA.Seattle'' that encapsulate multiple data in one value.

Copyright © 2005 Microsoft Corporation.
All rights reserved.