The xMart platform is a data harmonization and warehousing solution used by the World Health Organization (WHO) for health data. This tool contains a collection of features to help WHO data managers standardize, integrate and and store in a format that is ready for analysis.

xMart is a component of the World Health Data Hub (WHDH) and a publicly accessible information page about xMart and it role in the WHDH is available.

The tool is inspired by the need to combine and analyze files of various formats (over which you have no control), such as this one:

image-20210315155513512

Working with a file such as this can be very difficult, especially if your main job at the moment is to produce analytic outputs. Instead of doing analysis, you end up cleaning data.

xMart encourages a “datamart model”: prepare your team’s data once in a centralized repository rather than in an ad-hoc manner.

image-20210315154759063

xMart is organized as a collection of virtual “data marts” which physically reside in a single SQL Server database. Data marts can then be managed in a self-service manner by data mangers, including the ability to independently create tables, pipelines (connections to data sources), manage security, etc.

image-20210315154713298

Selected Screenshots of xMart

To provide an introductory understanding of the system, several screen shots are shown.

Top-level home page

Provides an overview of the system and all data marts.

img

Mart-level home page

For daily work, data managers of this mart would bookmark this page.

image-20210315165847769

Data model page

Users can create and change data models. This affects physical tables in the database.

img

Browsing data

Data tables or registered custom SQL views can be annotated, categorized (menu items) and tagged.

img

Viewing one data table

Includes basic features such as sorting, filtering and exporting.

img

Data upload page

Launch an upload from this page.

img

Preview data changes

Preview data before committing. System detects new, updated, unchanged and soft-deleted records.

img

View data issues during upload

img

List of ETL pipelines

Users can create pipelines in the web UI.

img

Pipeline editor

A pipeline is an online xml “file”. Includes autocomplete support, pre-run validation of pipeline and a debugging possibility.

img

Pipeline scheduling

img

Pipeline script guide

A pipeline command script guide is available providing “before” and “after” transformation examples.

img

Admin security page

Role-based security

img

OData API

A separate application exposes data as OData

img