1.what is incremental loading?
2.what is batch processing?
3.what is cross reference table?
4.what is aggregate fact table

Incremental loading means loading the ongoing changes in the OLTP.

Aggregate table contains the [measure] values ,aggregated /grouped/summed up  to some level of hirarchy.

let we explain these stuff like these... for eg you are working on universe it has n numbers of universe/classes/derived table.let we say these interms of objects.right now iam having 10 object(may be measure/dim/detail object what ever it may have).suppose i have created one object since now we have 11 object in the universe.if i export only the newly created object means it is called Incremental Loading or else when i export 10+1(newobjects) means that is normally export we doing always.

Incremental Loading:
Its also called retrospective data load.In this data being loaded on a regular inerval,and only new data is moved.
eg in ur data warehouse ur loading "Transactional Data" on daily basis and its happning only for new Transactions because old Data is already in the Data warehouse.
Batch Processing:
Its processing of Jobs/Task together.
eg if u have a staging area, u may have valid reason to get data for different Departments there first, and then move all that in one shot to production.

cross reference table used to describe many-2-many relation between tables.

aggregate fact table - when granularity of fact data not required fact can be grouped over dimensions, measures will be affected as sum(measure) or over aggregated functions.

Despite aggregation in the fact the drill-down possible into initial detail fact (raw fact) data using vice-verse of aggregation logic.