Papers by Zeqian Huang

1 papers
Typos Correction Training against Misspellings from Text-to-Text Transformers (2024.lrec-main)

Copied to clipboard

Challenge: Existing dense retrieval systems suffer from typoed queries due to mistyping or phonetic typing errors.
Approach: They propose a method that incorporates the spelling correction objective into the DR model and a prompt-based augmentation technique to enhance the alignment of the typoed query and its original query.
Outcome: The proposed model outperforms existing typos-aware training approaches and sophisticated training advanced retrievers.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations