Skip to main content

Quickstart - Create Semantic Data Fabric

This guide outlines the steps on how to use R!AN to create a semantic data fabric over your private data by initializing data sources and executing semantic data fabric queries.

Access the R!AN Dashboard​

Once you have started the R!AN server, you can access the R!AN dashboard to start creating semantic data fabric with your data.

Steps to Access the Dashboard​

  1. Start the R!AN Server

    Depending on your configuration, start the R!AN server using one of the following commands:

    With PostgreSQL:


    rian serve --config path/to/your/node.yaml

    With SurrealDB:


    rian serve

  2. Navigate to the Dashboard

Starting a R!AN server in serve mode will automatically open the dashboard in a browser tab.

r!an dashboard


Using the R!AN Dashboard​

Overview​

The R!AN dashboard provides an intuitive interface for managing your data, including sections for Cluster Info, Node Info, Discovery, Insights, Semantic Service, and Node Health.

Cluster Info​

Cluster Information: Provides details about the cluster, including nodes, status, and configuration.

Node Info​

Node Details: Displays information about the current node, including its ID, status, and configuration.

Discovery​

Service Discovery: Manages the data discovery services within the cluster.

Insights​

Insights Management: Allows you to view and manage insights generated by the system.

Semantic Service​

Semantic Pipelines: Runs semantic pipelines assigned by the control plane.

Node Health​

Health Monitoring: Monitors the health and performance of the node, providing detailed metrics and status reports.

Setup Data Sources​

r!an dashboard


Initialize Data Source(s)​

r!an dashboard


To initialize a local file folder source, use the following configuration:

{
        "backend": {
        "files": {
           "root_path": "path/to/your/local/file/folder",
           "id": "File Collector 1"
          }
      },
      "name": "Demo Collector"
    }

In this example, we are initializing a single data source. You can add additional data sources based on your specific data locations and requirements.

Create Semantic Data Fabric​

Start the Semantics service​

To start processing the data and create a semantic data fabric over your private data, follow these steps:

  1. Navigate to the Semantic Pipeline Request Semantic Data Fabric api

  1. Customize the Options
{
  "attention_threshold": 0,
  "collectors": [
    "string"
  ],
  "fixed_entities": {
    "entities": [
      "string"
    ]
  },
  "model": 0,
  "sample_entities": {
    "entities": [
      "string"
    ]
  }
}

  • attention_threshold: R!AN uses attention matrices from transformer models to construct the semantic data fabric. This threshold filters out data fabrics that fall below the specified score.
  • collectors: A list of sources to be used for fetching data.
  • fixed_entities: A list of entities around which the semantic data fabric will be constructed.
  • sample_entities: The types of the entities specified in fixed_entities.
  • model: Choose the model to use for Named Entity Recognition (NER).
    • Pass 0 to use the English-based NER model.
    • Pass 1 to use the Geology-based NER model.
    • If no model and no entities are specified, the default is the English-based NER model.
  1. Successful execution of the Semantic Pipeline Request will return a unique pipeline_id. This pipeline_id acts as a unique identifier for the semantic data fabric created during this session. This feature allows users to create different data fabrics using various semantic pipeline options, which is usef ul for :
  • Flexibility: Users can create multiple semantic data fabrics over the same data or different data sets. This flexibility allows for comparative analysis and tailored insights based on different configurations and parameters.
  • Customization: Each pipeline_id corresponds to a specific semantic pipeline configuration. Users can experiment with different settings to optimize their data discovery and insight generation processes.
  • Organized Data Management: With unique pipeline_ids, users can manage and organize their data fabrics efficiently, ensuring that each fabric can be referenced, modified, or analyzed independently.
  • Enhanced Insight Generation: By creating multiple semantic data fabrics, users can uncover a broader range of insights and patterns, improving decision-making and strategic planning.

Semantic Data Fabric api


Important Note

Remember to note down the pipeline_id returned after each successful execution of a semantic pipeline. This ID is crucial for performing searches and generating insights later. Each pipeline_id ensures that you can accurately reference and utilize the specific data fabric created during your session.

Summary​

The R!AN dashboard simplifies the management of your data and services, providing a user-friendly interface to interact with and monitor your cluster. By following the steps outlined in this guide, you have created a semantic data fabric using R!AN.