Bank Customer Transaction and Relevance Dataset

Columns

• user-id: The unique identification number assigned to the customer upon registration.
• user-age: The customer's age, ranging from approximately 21 to 74 years.
• user-gender: The customer's gender, categorised as male or female.
• user-nationality: Indicates if the customer is local or foreign.
• user-knowledge: The customer's level of investment knowledge (low, medium, or high).
• user-loyalty: A categorisation of purchasing consistency (e.g., new, sporadic).
• user-loan: Boolean value indicating whether the customer has a loan obligation.
• user-income: The customer's monthly income.
• user-savings: The total amount of savings held by the customer.
• user-properties: The number of properties owned by the customer.
• user-riskAversion: The customer's risk aversion level, classified as low or high.
• user-marital: Marital status (married, single, or other).
• user-dependents: The number of dependents relying on the customer.
• user-pension: The current value of the customer's pension.
• product-type: The category of product purchased (e.g., ETF, bond).
• product-risk: The risk level associated with the product (high, low, or other).
• product-term: The maturity duration of the product.
• product-yield: The yield level of the product (high, low, or other).
• transaction-id: A unique identifier for the specific transaction.
• year: The year the transaction occurred (ranging from 2015 to 2019).
• month: The month the transaction occurred.
• score: The target variable representing the relevance score of the transaction to the user (0 to 1).

Distribution

Usage

• Classification: Categorising customers based on their likelihood to purchase specific product types.
• Regression: Predicting the precise relevance score of a product for a new or existing customer.
• Customer Profiling: Analysing the relationship between demographic factors (age, income, dependents) and investment choices.
• Sales Optimisation: Targeting specific user segments (e.g., high savings, low risk aversion) with relevant product offers.

Coverage

• Geographic: Distinguishes between local (81%) and foreign (19%) customers.
• Time Range: Transactions cover the period from 2015 to 2019.
• Demographic: Includes a diverse age range (21-74), gender split (55% male, 45% female), and financial variation (e.g., income ranging from roughly 1,500 to over 24,000).

License

Who Can Use It

• Data Scientists: For training machine learning models to predict customer preferences.
• Financial Analysts: To understand market trends and customer investment behaviour.
• Marketing Teams: To segment customers and tailor promotional campaigns for financial products.
• Product Managers: To assess the performance and relevance of different investment products (ETFs, bonds) across user groups.

Dataset Name Suggestions

• Financial Product Relevance Prediction Data
• Customer Investment Profile and Recommendation Scores
• Bank Customer Transaction and Relevance Dataset
• Investment Product Suitability Score Data

Attributes