Credit Card Transaction Fraud Flags

This dataset provides detailed credit card transaction records enriched with fraud suspicion flags, risk scores, and contextual information such as merchant, location, and transaction method. It is ideal for developing, training, and evaluating fraud detection models, as well as for analyzing transaction patterns and identifying emerging fraud tactics in the financial sector.

  • opened 6 times
  • last updated 12 Jul 2025
  • by GoMask
The brief that made it

Training and benchmarking machine learning models for credit card fraud detection

Sample rows

preview · 8 of 150 rows · all 21 columns
transaction_idstringtransaction_typestringfraud_scorefloatcard_presentbooleantransaction_statestringcard_number_hashstringtransaction_datetimedatetimeamountfloatcurrencystringmerchant_idstringmerchant_namestringmerchant_category_codestringentry_modestringterminal_idstringtransaction_citystringtransaction_postal_codestringtransaction_countrystringis_internationalbooleanfraud_flagbooleanfraud_reason_codestringcustomer_id_hashstring
TXN001purchase0.03trueNYCARDHASH0012023-11-17T14:23:00Z42.5USDM001Whole Foods Market5411chipTERM1001New York10001USfalsefalseblankCUSTA
TXN002purchase0.08trueCACARDHASH0022024-01-05T09:11:47Z213.99USDM002Target5311contactlessTERM1002Los Angeles90001USfalsefalseblankCUSTB
TXN003purchase0.05trueWACARDHASH0032022-09-10T23:59:59Z9.95USDM003Starbucks5814swipeTERM1003Seattle98101USfalsefalseblankCUSTC
TXN004purchase0.04falseILCARDHASH0042023-07-07T12:00:00Z68.4USDM004Uber4121manualblankChicago60601USfalsefalseblankCUSTD
TXN005purchase0.07trueONCARDHASH0052022-12-25T18:15:12Z899CADM005Hudson's Bay5311chipTERM1004TorontoM5H1T1CAtruefalseblankCUSTE
TXN006purchase0.09trueblankCARDHASH0062023-03-30T21:47:33Z1542.75EURM006Galeries Lafayette5651chipTERM1201Paris75009FRtruefalseblankCUSTF
TXN007online0.02falseblankCARDHASH0072024-04-02T10:01:23Z5.99USDM007Apple App Store5818onlineblankblankblankUSfalsefalseblankCUSTG
TXN008purchase0.06trueTXCARDHASH0082023-10-01T08:12:04Z155USDM008Shell5541swipeTERM1005Houston77002USfalsefalseblankCUSTH

What the 150 rows show

from the 150-row sample

Purchase (transaction type) stands out: 99 of its 117 rows have card_present = true, against 5 of 33 for the rest.

  • 69%card_present = true
  • 0.05median fraud_score
  • 3fraud reason codes
  • 5entry modes
  • 11currencies
  • 13transaction countries
Card present rate by transaction_typecard_present = true
0%50%100%85%purchase99 of 1170%refund0 of 120%online0 of 16100%in-store5 of 5
fraud_score150 rows, in bands of 0.1
065130127200000002100.51fraud_score →

Median 0.05, from 0.01 to 1.0.

transaction_state105 rows with a value · 45 left blank
  1. CA22
  2. TX13
  3. FL8
  4. OH7
  5. CO6
  6. MA5
  7. NSW5
  8. NY4
  9. ON4
  10. WA3
21 columns by typefrom the column list below
  • string 15
  • float 2
  • datetime 1
  • boolean 3

Columns

21 columns in four groups
blueprint · 21 columns
columntypedescriptionexample
Text 15 columns
transaction_idstringUnique identifier for each credit card transaction.uniqueTXN001
card_number_hashstringHashed or tokenized credit card number to preserve privacy while enabling grouping by card.CARDHASH001
currencystringISO 4217 currency code used for the transaction (e.g., USD, EUR).11 currenciesUSD
merchant_idstringUnique identifier for the merchant processing the transaction.M001
merchant_namestringName of the merchant where the transaction occurred.optionalWhole Foods Market
merchant_category_codestringMerchant category code (MCC) indicating the type of business.optional5411
transaction_typestringType of transaction (e.g., purchase, refund, withdrawal, online, in-store).6 valuespurchase
entry_modestringHow the card was used (e.g., chip, swipe, contactless, manual entry, online).6 values · optionalchip
terminal_idstringIdentifier for the payment terminal used in the transaction.optionalTERM1001
transaction_citystringCity where the transaction took place.optionalNew York
transaction_statestringState or province where the transaction took place.optionalNY
transaction_postal_codestringPostal code of the transaction location.optional10001
transaction_countrystringCountry where the transaction occurred (ISO 3166-1 alpha-2 code).US
fraud_reason_codestringShort code or label indicating the reason for the fraud suspicion (e.g., 'velocity', 'location_mismatch', 'high_amount').3 codes · optionalhigh_amount
customer_id_hashstringHashed or tokenized unique identifier for the cardholder.optionalCUSTA
Numbers 2 columns
amountfloatTransaction amount in the local currency.0.01 or more42.5
fraud_scorefloatFraud risk score assigned by the detection model (0.0 to 1.0, higher means more likely fraudulent).0 to 1 · optional0.03
Dates and times 1 column
transaction_datetimedatetimeDate and time when the transaction occurred.2023-11-17T14:23:00Z
True or false 3 columns
is_internationalbooleanIndicates if the transaction is cross-border (true) or domestic (false).optionalfalse
fraud_flagbooleanIndicates if the transaction is suspected or confirmed as fraudulent.false
card_presentbooleanIndicates whether the physical card was present during the transaction.optionaltrue

Use it for

  • card present69%104 of 150 rowsmean fraud score by t…0.19purc…0.02refu…0.21onli…0.02in-s…

    A finance dashboard

    The card_present rate, fraud_score by transaction_type and a breakdown of transaction_state. Excel, Power BI or Tableau.

  • Why do 104 of 150 rows have card_present = true?

    A root-cause class exercise

    Hand out the rows and one question. The answer is in the data, not in the brief.

  • A software demo

    Believable transactions with card_number_hash, transaction_datetime and amount to fill a screen in front of a buyer.

Not quite right?

Make it yours.

Same 21 columns, your size and your rules. See 20 rows before you pay.

Preview 20 rows free

1,000 rows: $1.99 · 10,000: $12.99One-time. No subscription. All prices

This dataset150 rows21 columns
Yours1,000 rows21 columnstransaction_city: UK only

blueprint · credit-card-transaction-fraud-flags

Behind this dataset

Same schema. As many rows as you need.

These 150 rows came out of a blueprint — 21 columns with generation rules behind each one. Open it in Data Factory to retune a column, add your own, wire in foreign keys, and run it at the size you actually need.

Rules it was built with
  • Flag transactions exceeding $2000 as high risk
  • Include geolocation and merchant details
  • Classify transaction type (online, POS, ATM)
  • Record time and frequency per cardholder
  • Mark confirmed frauds for training labels
Rows
Open the blueprint in Data Factory

1 credit per row. Preview 20 rows free. Buy rows for the full run.

Exports
CSV, JSON, JSONL, Parquet, SQL, Excel, TSV, XML
Licence
yours to use, including commercially
API slug
credit-card-transaction-fraud-flags

More in Finance

Browse all
Also in FinanceFinancial Transaction Fraud PatternsTransaction Categorization DatasetAccounts Receivable and Aging AnalysisAll 1,954 datasets

What should your data show?

Preview 20 rows free
No signup. No card.