Extracting Spot Listing Pairs and Times from Exchange Announcements

Summary

This document presents a workflow for processing exchange announcements with a language model instead of relying on hand-written regular expressions. Given announcement text, the function asks the model to decide whether it announces new spot trading pairs and, when it does, return the pair symbols and listing time converted to Beijing time. The example covers an announcement with multiple pairs and distinguishes spot listings from other details such as margin pairs and withdrawals.

The approach can handle varied announcement wording and time formats, but the document gives no measured accuracy or comparison with rule-based parsing. Its output depends on the model following the requested JSON format, and the shown implementation parses the response directly. Incorrect extraction, timezone conversion, or malformed output could affect downstream use, so results need validation before they inform an event-driven trading workflow.

Key ideas

• The workflow uses a language model to classify whether an announcement reports new spot pairs.
• For qualifying announcements, it extracts pair symbols and converts the listing time to Beijing time.
• The example distinguishes spot listings from margin listings and withdrawal schedules.
• The document reports no accuracy testing, and the returned data requires validation.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.