[]
(Showing Draft Content)

Vectorized Data

The Vectorized Data section enables administrators to manage different entities: Datasets, Data Model, and User Knowledge that have been vectorized for AI-powered semantic search and embedding models. This functionality is essential for improving data retrieval and enabling natural language query performance.


Vectorized Data

Dataset

The Dataset section under Vectorized Data enables administrators to manage datasets that have been vectorized for AI-powered semantic search and embedding models.

Navigate to: System Management > Vectorized Data > Dataset.

Options:

  • Only show records without vectorized data: Filter datasets that have not yet been vectorized.

  • Search Bar: Quickly locate datasets by name.

  • Batch refresh vectorized data: Re-run vectorization for multiple datasets at once.

  • Title: Name of the dataset (e.g., HealthDataSet: National, FinanceDataset2, CustomerOrders).

  • Creator: User who created the dataset (e.g., admin).

  • Last Update: Timestamp of the most recent update.

  • LLM: Indicates the embedding model used (e.g., text-embedding-v2).

  • State: Current status of vectorization (e.g., Available).

    • After vectorization, the State column should display Available.

    • The LLM column confirms the embedding model applied.

Data Model

The Data Model section under Vectorized Data enables administrators to manage data models that have been vectorized for AI-powered semantic search and embedding models.

Navigate to: System Management > Vectorized Data > Data Model.

Options:

  • Only show records without vectorized data: Filter data models that have not yet been vectorized.

  • Search Bar: Quickly locate data models by name.

  • Batch refresh vectorized data: Re-run vectorization for multiple data models at once.

  • Title: Name of the data model (e.g., Demo 1, Demo).

  • Creator: The user who created the data model.

  • Last Update: Timestamp of the most recent update.

  • LLM: Indicates the embedding model used.

  • State: Current status of vectorization (e.g., Available, Missing).

    • After successful vectorization, the State column should display Available.

    • If incomplete, missing may appear.

User Knowledge

The User Knowledge section under Vectorized Data enables administrators to manage knowledge bases that have been vectorized for semantic search and AI-driven insights.

Navigate to: System Management > Vectorized Data > User Knowledge

Options:

  • Only show records without vectorized data: Filter knowledge bases that have not yet been vectorized.

  • Search Bar: Quickly locate knowledge bases by name.

  • Batch refresh vectorized dataRe-run vectorization for multiple knowledge bases at once.

  • Title: Name of the knowledge base.

  • Creator: The user who created the knowledge base.

  • Last Update: Timestamp of the most recent update.

  • LLM: Indicates the embedding model used.

  • State: Current status of vectorization (e.g., Available, Missing).