Explore the cutting-edge developments in natural language processing and its application to scientific research acceleration in this 44-minute conference talk by Daniel Weld at KDD. Dive into the world of Semantic Scholar, a powerful AI-driven platform revolutionizing academic literature search and discovery. Learn about longform language models, adaptive research feeds, and personalized recommendation systems. Discover innovative features like extreme summaries, information nuggets, and the ScholarPHI system. Gain insights into the research journey behind these advancements, including user studies and the development of the Semantic Reader. Understand how these tools are enhancing the efficiency and effectiveness of scientific research, and get a glimpse of future possibilities in AI-assisted academic exploration.
Semantic Scholar- Tera-Scale NLP to Accelerate Scientific Research--Daniel Weld
Association for Computing Machinery (ACM) via YouTube
Overview
Syllabus
Semantic Scholar
About Semantic Scholar
Longform Language Model
Language Models
Data Resources
White House Contact
Research Journey
Adaptive Research Feed
Research Dashboard
Building a recommendation system
Recommendation based on content
Training data
Citation graphs
Inference
Visualization
Personalized Alerts
Clickthrough Rate
Finding Papers
Semantic Scholar Features
Extreme Summary
TLRs
How good
How informative
Information Nuggets
Accuracy
Initial studies
ScholarPHI system
Demo
User Study
Semantic Reader
Foresight
Summary
Supported Decoding
Restricted Decoding
Roberta
Recap
Thanks
Interview
Taught by
Association for Computing Machinery (ACM)
