HiDy: A Large-scale Hierarchical Dynamic Financial Knowledge Base

HiDy is a hierarchical, dynamic, robust, diverse, and large-scale financial KB that aims to provide various valuable financial knowledge as critical benchmarking data for fair model testing in different financial tasks. Specifically, HiDy currently contains 34 relation types, more than 505,821 relations, 17 entity types, and more than 51,095 entities. The scale of HiDy is steadily growing due to its continuous updates. To make HiDy easily accessible and retrieved, HiDy is organized in a well-formed financial hierarchy with four branches, Macro, Meso, Micro, and Others.

We then give explanations on the various csv files as follows.

• "hidy.nodes.entity_type.csv" includes a mapping dictionary of entities with a specific entity type. The meta data is ID, name, (code), Label. For example, "hidy.nodes.company.csv" includes "0,东诚药业,002675.SZ,company", "1,大庆华科,000985.SZ,company".
• "hidy.relationships.relation_type.csv" includes quadruple knowledge with a specific relation type. The meta data is START_ID, END_ID, TYPE, time. For example, "hidy.relationships.cooperate.csv" includes "2197,245,cooperate,2019/9/17 12:00, "1165,756,cooperate,2020/8/2 17:10".

For details of HiDy, please refer to our GitHub.