One or more working objects (106), e.g., email messages and texts, are categorized into one or more of a plurality of categories (112A, 112B) based on the content/attributes of each working object versus the content/attributes of objects already categorized. A category profile (154A, 154B) is created for each category from a plurality of training objects (142) known to be categorized under that category. An object profile (124C) is created for each working object. Each working object profile is then compared to each of the category profiles to determine information regarding goodness-of-fit of that object profile to each of the category profiles. The corresponding working object is categorized based on this goodness-of-fit information.