Skip to main content

🔍 Dimension Members Retriever

In the meta-analysis platform, the Dimension Members Retriever is a key feature within the Copilot Command Agents. It provides more precise and efficient support for data analysis by dynamically synchronizing and quantifying dimension member information.

Synchronize Dimension Members and Vectorize Information​

The Dimension Members Retriever synchronizes dimension members on the member retrieval page of the semantic model and processes them into vectors. Specifically, it can:

  • Dynamically synchronize dimension members: Sync dimension member information from databases or other data sources. This means that whenever the data source changes, the dimension member information is automatically updated, ensuring the timeliness and accuracy of the analysis data.

  • Quantify dimension member information: During synchronization, the backend processes the dimension member information into vectors and saves these vectors in the dimension member table.

Dimension Member Retriever
Dimension Member Retriever

Practical Use Cases​

1. Dynamic Retrieval during Copilot Command Execution​

When a user issues data analysis commands through Copilot Command, the Dimension Members Retriever dynamically retrieves relevant dimension member information based on the specific prompt information. The specific steps are as follows:

  • Parse user commands: Copilot Command first parses the user input commands and prompt information to determine the required analysis tasks and related dimensions.

  • Invoke the Dimension Members Retriever: Based on the parsing results, the Agent invokes the Dimension Members Retriever to retrieve dimension member information relevant to the current analysis task. For example, if the user needs to analyze sales data for a particular product category, the Retriever will automatically retrieve and return detailed information for all products under that category.

  • Integrated information processing: The information returned by the Dimension Members Retriever includes quantized data and key information of all relevant dimension members, laying the foundation for subsequent data processing and analysis.

2. Integrated Information and LLM Collaboration​

After obtaining the quantized information of the dimension members, the Agent sends this information to the large language model (LLM) and processes it further in conjunction with the user's prompt information. The specific process is as follows:

  • Send and integrate information: The Agent sends the quantized information of the dimension members and the user's prompt to the LLM. The LLM comprehends this information to understand the user's specific needs and analysis goals.

  • Invoke tools to perform operations: Based on the dimension member information and the prompt, the LLM invokes the appropriate analysis tools to perform operations. These tools may include data filtering, statistical analysis, visualization, and more, ensuring that the final analysis results meet the user's expectations.