Can we identify duplicate accounts before importing data from a CSV file?

Duplicate Identification can't run against a CSV file before importing the data in the file.

Duplicate Identification batches run only on data in Oracle Fusion applications. The duplicate identification feature works on composite key generation settings that are part of the persisted data model and aren't present or calculated for data in the CSV during the import process.