Better Extraction from Text Towards Enhanced Retrieval (BETTER)

The Better Extraction from Text Towards Enhanced Retrieval (BETTER) Program will develop methods for extracting increasingly fine-grained semantic information, with a focus of events in the form of who-did-what-to-whom-whenwhere, across multiple languages and problem domains. This extracted information will be applied to an information retrieval task. An additional area of focus is human-in-the-loop computation. Performer systems will need the ability to incorporate human judgments for metrics such as relevancy and accuracy of extracted or retrieved information.

The UMD focused on improving representations for low resource langauges through using related languages and human interaction.

	Jordan Boyd-Graber Assistant Professor, Computer Science (UMD)
	Mozhi Zhang PhD Student, Computer Science (UMD)

<< back to top

• Mozhi Zhang and Yoshinari Fujinuma and Michael J. Paul and Jordan Boyd-Graber. Why Overfitting Isn't Always Bad: Retrofitting Cross-Lingual Word Embeddings to Dictionaries. Association for Computational Linguistics, 2020. [Preprint] [Video] [Code] [Bibtex]
  
  

  @inproceedings{Boyd-Graber-2020,
  	Title = {Why Overfitting Isn't Always Bad: Retrofitting Cross-Lingual Word Embeddings to Dictionaries},
  	Booktitle = {Association for Computational Linguistics},
  	Year = {2020},
  	Location = {The Cyberverse Simulacrum of Seattle},
  	Url = {http://umiacs.umd.edu/~jbg//docs/2020_acl_refine.pdf}
  }
  

  
  

Acknowledgments

Any opinions, findings, and conclusions or recommendations expressed in this material are those of the researchers and do not necessarily reflect the views of the sponsor.