You are given a dataset of credit card transaction records and suspect that some records are duplicates. Discuss: - What real-world situations could create duplicate-looking records. - How you would define a true duplicate, especially when two records are very similar but not exactly identical. - How you would detect, quantify, and extract duplicate records from the dataset in a robust way. Your answer should consider both exact duplicates and near-duplicates caused by ingestion issues, retries, formatting differences, or multiple processing stages.