Abstract
The technology described herein is directed to a method for accurately classifying user intent from highly ambiguous text strings. This can be used to classify online advertising deal names that require Special Category Data (SCD) compliance. As deal names inherently lack context, frequently use acronyms, and have minimal metadata associated with them, it is difficult to interpret the deal names for SCD relevance. This is addressed by using artificial intelligence (“AI”) models that can expand/interpret the deal names in a first phase and use the expansion/interpretation in a second phase to classify the deal name relevance for SCD purposes.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Stewart, Chris, "Expansion-Classification Prompt with “Thinking” Module for the Interpretation of Highly Ambiguous Inputs", Technical Disclosure Commons, (August 19, 2025)
https://www.tdcommons.org/dpubs_series/8484