Abstract
A system for determining semantic similarity of phrases (word sequences) uses weak supervision of neural networks to generate an embedding space for determining phrase similarity. The training of the neural networks uses input from two orthogonal sources: click distribution of queries and phrase translations.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Uszkoreit, Jakob, "Identifying Similar Meaning In Word Sequences", Technical Disclosure Commons, (April 24, 2017)
https://www.tdcommons.org/dpubs_series/480
COinS