Better Extraction from Text Towards Enhanced Retrieval (BETTER)

Project funded by IARPA
PI: Jordan Boyd-Graber

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

• Yoshinari Fujinuma, Jordan Boyd-Graber, and Katharina Kann. How Does Multilingual Pretraining Affect Cross-Lingual Transferability?. Association for Computational Linguistics, 2022. [Code] [Bibtex]
  
  

  @inproceedings{Fujinuma:Boyd-Graber:Kann-2022,
  	Author = {Yoshinari Fujinuma and Jordan Boyd-Graber and Katharina Kann},
  	Title = {How Does Multilingual Pretraining Affect Cross-Lingual Transferability?},
  	Booktitle = {Association for Computational Linguistics},
  	Year = {2022},
  	Location = {Dublin},
  	Url = {http://umiacs.umd.edu/~jbg//docs/2022_acl_multilingbert.pdf},
  }
  

• Mozhi Zhang, Yoshinari Fujinuma, 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{Zhang:Fujinuma:Paul:Boyd-Graber-2020,
  	Author = {Mozhi Zhang and Yoshinari Fujinuma and Michael J. Paul and Jordan Boyd-Graber},
  	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},
  }
  

  
  
• Michelle Yuan, Hsuan-Tien Lin, and Jordan Boyd-Graber. Cold-start Active Learning through Self-Supervised Language Modeling. Empirical Methods in Natural Language Processing, 2020. [Video] [Code] [Bibtex]
  
  

  @inproceedings{Yuan:Lin:Boyd-Graber-2020,
  	Title = {Cold-start Active Learning through Self-Supervised Language Modeling},
  	Author = {Michelle Yuan and Hsuan-Tien Lin and Jordan Boyd-Graber},
  	Booktitle = {Empirical Methods in Natural Language Processing},
  	Year = {2020},
  	Location = {The Cyberverse Simulacrum of Punta Cana, Dominican Republic},
  	Url = {http://umiacs.umd.edu/~jbg//docs/2020_emnlp_alps.pdf},
  }
  

  
  
• Michelle Yuan, Mozhi Zhang, Benjamin Van Durme, Leah Findlater, and Jordan Boyd-Graber. Interactive Refinement of Cross-Lingual Word Embeddings. Empirical Methods in Natural Language Processing, 2020. [Audio of Paper Readthrough] [Video of Paper Readthrough] [Code] [Conference Talk] [Bibtex]
  
  

  @inproceedings{Yuan:Zhang:Van-Durme:Findlater:Boyd-Graber-2020,
  	Title = {Interactive Refinement of Cross-Lingual Word Embeddings},
  	Author = {Michelle Yuan and Mozhi Zhang and Benjamin {Van Durme} and Leah Findlater and Jordan Boyd-Graber},
  	Booktitle = {Empirical Methods in Natural Language Processing},
  	Year = {2020},
  	Location = {The Cyberverse Simulacrum of Punta Cana, Dominican Republic},
  	Url = {http://umiacs.umd.edu/~jbg//docs/2020_emnlp_clime.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.