Scholarlink

This article explains how Model Driven Engineering (MDE) and Data Warehouses (DWH) work together through the ideas of abstraction and transformation. MDE, created by the Object Management Group (OMG), uses different model levels (CIM, PIM, PSM, and Code) and transformations to keep models clear and consistent. Standards like QVT and ATL make these transformations reliable. At the same time, Data Warehouses collect and organize data to support Business Intelligence (BI). They use facts and dimensions to analyse data in many ways with OLAP tools. The Semantic Layer adds a shared business language and gives a single, clear view of the data. Our article proposes an LLM-driven method to transform a source model into a target model oriented toward the Data Warehouse. This method automatically generates the semantic layer, linking the technical model to the business meaning of the data. It shows how MDE, LLM, and BI can work together to create intelligent transformation process.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

LLM-Based Semantic Layers: Connecting Model-Driven Engineering and Data Warehouses

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Ibtissam Arrassen,
  • Youssef Ben Amar,
  • Mohammed Erramdani

摘要

This article explains how Model Driven Engineering (MDE) and Data Warehouses (DWH) work together through the ideas of abstraction and transformation. MDE, created by the Object Management Group (OMG), uses different model levels (CIM, PIM, PSM, and Code) and transformations to keep models clear and consistent. Standards like QVT and ATL make these transformations reliable. At the same time, Data Warehouses collect and organize data to support Business Intelligence (BI). They use facts and dimensions to analyse data in many ways with OLAP tools. The Semantic Layer adds a shared business language and gives a single, clear view of the data. Our article proposes an LLM-driven method to transform a source model into a target model oriented toward the Data Warehouse. This method automatically generates the semantic layer, linking the technical model to the business meaning of the data. It shows how MDE, LLM, and BI can work together to create intelligent transformation process.