It's better to think about the idea of a map rather than all the different kinds. A map has information, usually the values are stored as gray scale values. So one might be white and zero black. This can be put into a huge number of attributes to affect a mesh differently over its surface. For instance, a specular map will change which areas are more or less specular. If you know what an attribute does, then you know what a map of that attribute will do.
Focus on the attribute and what it does instead of trying to memorize all possible map names. Why?
A map is just an image, the value of each pixel provides a value for the attribute at the matching point on the object. Because there are some attributes/channels that are so commonly mapped, it resulted in maps being named after them (color/diffuse, transparency, specular, bump, ect) but trying to remember all the different types of maps is like trying to remember all the attributes that can be mapped, it's impossible. You can map even user created attributes. To expand on what Stwert said, if I created a jiggle attribute that takes a value of 0 to 1 (0 = no jiggle, 1 = full jiggle) then a grayscale image I generate to plug into it can safely be called a jiggle map, wherever the dark areas of my image lands on the object will have less jiggle than the light areas.