About semantic models
In this section, we provide an overview of semantic model; an essential piece of Sightfull; transforming raw data into meaningful information and creating a solid, consistent foundation for the metric layer. This introduction presents the what, why, how of semantic model and sets the stage for deeper exploration into its key components: entities, dimensions, and relationships.
What is semantic model?
Semantic model in the context of data analytics is the process of adding meaning to raw data by mapping it to objects, concepts and processes in the real world. It's about transforming data into a format that reflects the real-world scenarios and contexts your organization operates in.
Semantic Mapping is the process of organizing, categorizing, and relating data to create semantic models that mirror an organization's business logic and reality.
This process involves defining entities (like customers or products), their dimensions (such as age, category), and the relationships between them. Semantic model creates a layer of abstraction over raw data tables and columns, enabling users to interact with data in a more intuitive, focused and relevant manner.
For example, simple semantic mapping can map the accounts table to account entity and the raw table columns to the dimensions and relationships that are most important to us; such as the current account status, the date they became customers, all their relevant contacts and so on. We can also add default values, define logic for custom dimension calculation, create custom relationships to other entities and more.
Why is semantic model so useful?
Data has a tendency to get messy. When data is inconsistent or incomplete, it becomes incredibly difficult to trust and understand it enough to support even basic analytics. The semantic model solves for many such issues by applying logical operations to map data to a more meaningful format and automatically surface or resolve inconsistencies and gaps (for example by defining a default value for a dimension).
In this way, semantic model creates a flexible and robust foundation for the metric layer and all other subsequent data exploration and manipulation. The key benefits of using this semantic foundation are:
- Clarity: It's easier to understand what data means and what are it's key features.
- Consistency: Critical data is more consistent and complete when transformed by the logic and heuristics encoded in semantic models.
- Flexibility: Data can be adapted to changes in workflow or data source by updating the semantic models.
How can teams best leverage semantic models?
Sightfull utilizes semantic mapping to convert complex, often siloed data into unified, coherent semantic models. This unified logic is often referred to as an ontology (Read more about Sightfull's approach to data ontology here). Teams can leverage this comprehensive ontology and semantic mapping to:
- Embed business logic in their data: Capture and implement your business logic in your data by writing it into your semantic models.
- Break down silos: Map data from different sources to a unified data model using different semantic models as "translators".
- Align everyone on the same page: More meaningful, consistent and complete data creates the foundation for a shared view of what is happening.
- Increase clarity and actionability: With data presented in a business context, it becomes easier for analysts and decision-makers at all levels to understand and act on it.
Exploring key components of the semantic model
To understand how semantic mapping works in Sightfull, we’ll explore its key components:
- Entities: Learn about the building blocks of the semantic model – the representation of real-world objects or concepts.
- Dimensions: Discover how dimensions provide context and attributes to entities, enriching the data model.
- Relationships: Understand how relationships connect entities, enabling complex data associations and insights.
Wrapping Up
Semantic mapping is a transformative feature of Sightfull, turning raw data into a structured, meaningful, and business-aligned semantic model. It lays the foundation for creating a powerful and flexible metric layer that truly translates data to the unique real-world narrative of your organization.