Integrating data into the system is key to making your machine learning models work effectively. This process involves setting up a data source that can be automated once configured. All operations occur at the dataset level, so understanding how to create and manage datasets and data sources is crucial.

​
Data Mapping: Entities Description

Once you have a preview of the data, you need to map it to the allyy data structure, which organizes contact and interaction data.

​
Revised Terminology for Clarity:

  • Contacts: Individuals or entities that will interact with your system.
  • Offers/ContentItem/Treatment: Objects or items that a contact can interact with. Examples include Offers (e.g., promotions), ContentItems (e.g., articles), and Treatments (e.g., telemarketing).
  • Responses: Captures interactions between a contact and an object. This could include positive actions (e.g., clicks, donations) or negative actions (e.g., unsubscribes).
  • Subscriptions: Represents agreements between a contact and an object (e.g., subscriptions to services). They can have start and end dates, or be ongoing if the end date is undefined.
  • Lists: Collections of contacts or objects used in models or workflows (e.g., a list of target customers for a prediction).
Entities mapping diagram

​
Data Mapping Fields

When mapping your data to the allyy data structure, there are different categories of fields:
  • IDs: Unique identifiers for contacts, offers, or responses (e.g., ContactID, OfferID).
  • Attributes: Characteristics of the data (e.g., age, gender, postal code).
  • Classes and Types: Fields that help categorize data (e.g., Offer Type, Content Category).
  • Valid From: Indicates when the data is valid from, useful for tracking changes over time.
  • Mandatory Fields: Required fields that must be mapped (e.g., Subscription Start Date).
  • Optional Fields: Enrich the entity with additional information but are not required (e.g., Subscription Frequency).