Knowledge Base Article

Unify: What the workload does

This page provides a primer on Entity Resolution and the process of the Quantexa Unify workload for Microsoft Fabric.

Table of Contents

What is the Quantexa Unify workload?

The Quantexa Unify workload is an end-to-end, AI-driven Entity Resolution feature in the Microsoft Fabric ecosystem. The workload is built out of the Entity Resolution component of Quantexa’s industry-leading Decision Intelligence (DI) Platform.

For a more detailed overview of the Unify workload, see Unify: How the workload can help you.

What is Entity Resolution and why is it useful?

Entity Resolution is the process of working out whether multiple records are referencing the same real-world object, such as a person, organization, address, phone number, bank account, or device. The process takes multiple disparate data points from external and internal sources and resolves them into one distinct, unique Entity.

Ultimately, Entity Resolution cleanses, distils, and unifies your data, making it more accurate and useful. This improves data analysis and real-world decision-making and, with Quantexa’s best-in-class Entity Resolution capabilities, helps you unlock deeper insights and make smarter decisions with ease.

To learn more about Entity Resolution, and why it’s essential for businesses working with disparate datasets, see the following:

To learn more about why Entity Resolution is a much more effective way of handling disparate datasets compared to Record-to-Record matching, see the video below:

 

How will Entity Resolution through the Unify workload help me?

The Unify workload brings Quantexa’s industry-leading Entity Resolution capabilities to your doorstep in a secure and easy-to-use workload.

Providing fast data onboarding and a no-code user interface, the workload is straightforward and quick to use, enabling you to resolve Entities end-to-end in under an hour and eliminating the need for more complex tooling and software.

What does the Unify workload process look like?

The Quantexa Unify workload supports end-to-end Entity Resolution, from ingesting, mapping, and resolving your source data, to generating reliable outputs for more effective decision-making.

The process works as follows:

Connect your Data Source

First, you must connect your Data Source to a project. You can do so in just a few clicks.

Data Mapping

On connecting your Data Source, Unify automatically processes it to identify individual fields within the data. This stage is called Data Mapping

The Data Mapping process uses an inference engine to determine the appropriate data schema. It pulls fields it identifies from your Data Source and creates a list of these fields.

Each field is then mapped to predefined Entity Types. For example, the field DateOfBirth maps to the Individual Entity Type. You can also create new Entity Types for the Data Mapping stage, as needed.

This stage applies the necessary parsing, cleansing, and standardization. After it is complete, you can view both the input and output data using the Data Mapping features.

Run an Iteration

With just a few clicks, you can then run an Iteration with the Data Sources you want to resolve and specifying the Matching Level you want to apply. The Iteration stage is the core Entity Resolution stage.

Once you have started an Iteration, the workload automatically completes multiple steps as follows:

  1. The Entity Resolution process takes the parsed, cleansed, and standardized data from the Data Mapping stage and identifies connections between Entity occurrences.
    • Unlike traditional record-matching approaches, Unify’s process does not rely solely on matching unique identifiers. With Quantexa’s best-in-class Entity Resolution capabilities, Unify helps you uncover connections from indirect relationships too.
  1. On the basis of the Entity occurrence matching, the workload outputs cleansed, unified, and organized information about each Entity.
    • Quantexa’s proprietary code ensures that the output provides a more accurate, real-world representation of each Entity than your input data or traditional record-matching approaches.
  1. The workload then places its output tables into a Lakehouse of your choosing.
  2. In addition, it creates a Power BI report and a Semantic Model of the Iteration’s output tables.

NOTE: Although the workload creates the Semantic Model automatically, you are able to amend it.

Output and next steps

Following these steps, you have now resolved Entities and can use the output information to create more detailed reports, such as a Power BI report.

What output does the workload produce?

For each Iteration, the workload produces the following outputs automatically:

  • Entity Resolution records and Entities tables
  • Default Semantic Model
  • Unify Iteration summary
  • Power BI Report

For further details on these outputs, see Unify: A closer look at selected key features.

Additionally, using the automatic outputs, you can optionally create outputs within the broader Fabric suite, including the following:

  • Other types of Power BI report
  • Notebooks

Who can use Unify?

The Unify workload is particularly useful for Data Warehouse specialists, Business Analysts, Data Engineers, and Data Scientists.

However, due to the simplicity and no-code nature of Unify’s user interface, anyone working with uncleansed and disparate datasets can use Unify. Depending on the size of your dataset, you could run the Unify workload end-to-end in under an hour, giving you fast, accurate, and reliable Entity Resolution in just a few clicks.

Additionally, with a Fabric subscription, you can quickly onboard and start using the Unify workload anytime.

Next steps

See the following guides for further information on core concepts in Unify and guides to setting up and using your workload:

Updated 3 days ago
No CommentsBe the first to comment
Related Content