Darsh J Shah
Scholar

Darsh J Shah

Google Scholar ID: XscFiX8AAAAJ
Massachusetts Institute of Technology
Natural Language ProcessingMachine Learning
Citations & Impact
All-time
Citations
1,067
 
H-index
11
 
i10-index
11
 
Publications
15
 
Co-authors
33
list available
Resume (English only)
Academic Achievements
  • Publications: 'Generating Related Work for Scholarly Papers' (Preprint 2021), 'Nutri-bullets Hybrid: Consensual Multi-document Summarization' (NAACL 2021), 'Nutri-bullets: Summarizing Health Studies by Composing Segments' (AAAI 2021), 'Are We Safe Yet? The Limitations of Distributional Features For Fake News Detection' (Computational Linguistics 2020), 'Automatic Fact-guided Sentence Modification' (AAAI 2020), 'Capturing Greater Context for Question Generation' (AAAI 2020), 'Towards debiasing fact verification models' (EMNLP 2019), 'Robust Zero-Shot Cross-Domain Slot Filling with Example Values' (ACL 2019), 'Multi-Source Domain Adaptation with Mixture of Experts' (EMNLP 2018), 'Adversarial Domain Adaptation for Duplicate Question Detection' (EMNLP 2018).
Research Experience
  • Research projects include: Generating related work sections for scientific papers, generating consensual multi-document summaries, fact-guided modifications in text, etc.
Education
  • PhD: Massachusetts Institute of Technology (MIT), Electrical Engineering and Computer Science, Advisor: Prof. Regina Barzilay; B.Tech: Indian Institute of Technology Bombay (IIT Bombay), Computer Science and Engineering, Graduated in 2016.
Background
  • Research Interests: Natural Language Processing, Contrastive Text Generation, Scaling Machine Learning Models. Brief Introduction: PhD in Electrical Engineering and Computer Science from MIT, focusing on modeling the generation of contrastive text.
Miscellany
  • Contact: darsh@csail.mit.edu