Regardless of choosing the data warehousing methodology your frond end using Analysis Services would ultimately be a data mart. In simple a data mart can be understand as is a database that is organized according to the methodology you have chosen to your warehouse and is composed of fact and dimension tables.
The structure of the data mart separate the entire database into to two distinct entities.
Dimension table hold attributes of different entities that describe fact records that are part of fact tables, among these some are representing descriptive information while rest of the other are used to specify how attributes of fact table should be modeled and sketched that would be helpful to the analyst. Dimension table normally consists of hierarchies of attributes that support summarization phase. Let us consider an example of a dimension containing customer information that contain a hierarchy that divides customer in different categories and subcategories until it reaches at the lowest level.
Normally a data warehouse contain one or more fact table in their data mart structure, a fact table confined of measures, fact table normally contain huge number of rows as they contain historical data for number of years. To acquire the information about different part of the organization the fact table contain the numeric(fact) that are summarized to fulfill the task to analyze.
Fact tables should not hold descriptive information or any data other than the numerical measurement fields and the index fields that relate the facts to corresponding entries in the dimension tables.
The fact table is the composition of foreign and primary keys that are related to relevant dimension table.