User attributes

Auxia's typical engagement removes all the heavy lifting that comes with the data integration. Your teams are only required to share your relevant tables and our internal deployment teams will do all the transformation required to ingest this data into Auxia's system.

Background

User attributes are characteristics or qualities that define and describe a user. These attributes are crucial for personalizing the user experience and enabling targeted treatment recommendations to enhance the overall model impact.

Some common user attributes include:

• Demographic information: Age, gender, location, income, education level, etc.

• Behavioral information: App usage history, frequency of use, preferred features, etc.

• Psychographic information: Personality traits, interests, values, etc.

• Technical information: Device type, operating system, browser, etc.

User attributes can be collected through a variety of methods, such as surveys, questionnaires, and app analytics.

How Auxia uses user attributes

User attributes can be leveraged in several ways to improve the user experience:

• Qualification Criteria: Define rules to determine which users should receive a treatment and which should not.

• Content Personalization: Tailor treatment content by dynamically inserting specific attributes into the messaging.

• Enhanced Treatment Selection: Use user attributes to guide the model in selecting the most appropriate treatment for the user.

• Machine Learning Features: Features are measurable characteristics of data that are used by the model to make predictions. Think of these as the input variables that help the model understand and learn patterns in the data.

How to structure your attributes for ingestion

User attribute schema

User attributes can be structured in two main ways within a database schema to prepare them for ingestion into Auxia's system:

Type 1: Attributes in columns

In this approach, each user attribute is represented as a column in a table. For example, a table of users might have columns for user_id, name, age, gender, location.

user_id	name	age	gender	location	ocupation	user_type	device_type	app_version
1	John	30	Male	New York	Doctor	Frequent Purchaser	Android	1.2.3
2	Mary	25	Female	Boston	Teacher	Occasional Browser	iOS	2.0.1
3	Bob	40	Male	Chicago	Engineer	High-Value Customer	Windows Phone	1.0.0

Type 2: Attributes in rows

In this approach, each user attribute is represented as a separate row in a table. The attribute name is stored in one column, the attribute value is stored in another column and attribute type is another column. For example, a table of user attributes might have columns for user_id, attribute_name and attribute_value.

user_id	attribute_name	attribute_value	type
1	name	John	STRING
1	age	30	INTEGER
1	gender	Male	STRING
1	location	New York	STRING
2	name	Mary	STRING
2	age	25	INTEGER
2	gender	Female	STRING
2	location	Boston	STRING

Ingesting attributes into Auxia

User attributes ingested into Auxia have the following structure. Once your data is structured in this format, please complete the remaining steps here and contact your Auxia POC to complete the process.

[required] user_id: Unique identifier for every user

[required] user_properties: A list of key value pair which adds more context to the state of the user, such as, subscription_status, country, sign_up_date, life_time_value etc.

[required] update_timestamp: The timestamp at which the attributes were updated