Project Objective: Establish a macroeconomic database and AI economist to enhance internal macro investment research efficiency and provide automated macroeconomic interpretation to the public.
1. Macroeconomic Database: A proprietary macroeconomic database owned by the company, covering data categories such as China's macroeconomics, regional macroeconomics, global macroeconomics, and industrial economics.
Phase One: Covers over 6,000 main indicators of China's macroeconomics from the National Bureau of Statistics, researched and established a data model for the macroeconomic indicator system, and conducted multi-source data crawling, ETL, integration, quality inspection, and launching.
Phase Two: Covers the full 65,000+ indicators of China's macroeconomics from the National Bureau of Statistics, and more than 4,000 regional and international indicators. Completed the extraction of indicator dimensions and other metadata, defined the processing functions and procedures for derived/composite indicators.
Challenges and Solutions: The macroeconomic indicator system is complex, with a large number of indicators and variable methods of derived measurements. The National Bureau of Statistics used natural language to describe the indicators, which is hard for machines to understand. Through the governance of the indicator system, basic indicators, dimensions, derived measurements, units, frequencies, base periods, regions, and other information were extracted, maintained, and associated with related dimension tables, laying the foundation for automated indicator extraction and analysis.
2. Macroeconomic Analysis Dashboard: Targeted at government and corporate senior decision-making departments, it provides analysis, interpretation, and forecasts of global, Chinese, and even regional macroeconomics. Based on the macroeconomic analysis framework provided by researchers, indicators are selected, visualized, and automatic interpretations of indicator changes are performed.
Results: Established a macroeconomic dashboard, serving multiple collaborative projects.
