Semantic Scholar | Semantic Reader

Introducing Semantic Reader

An AI-Powered Augmented Scientific Reading Application

Background

Semantic Reader is an augmented reader with the potential to revolutionize scientific reading by making it more accessible and richly contextual.

Studies of scientists reading technical papers show that readers are subject to many points of friction that break the flow of paper comprehension:

Frequently paging back and forth looking for the details of cited papers
Challenges recognizing the same work across multiple papers
Losing track of reading history and notes
Contending with a PDF format that is not well suited to mobile reading or assistive technologies such as screen readers

Semantic Reader uses artificial intelligence to understand a document’s structure and merge it with the Semantic Scholar’s academic corpus, providing detailed information in context via tooltips and other overlays. For readers that log into Semantic Scholar, Semantic Reader integrates with your library and, over time, will incorporate personalized contextual augmentations as well.

Now Available

Semantic Reader is now available for most arXiv papers on semanticscholar.org with an introductory set of features.

Citations Cards that show details of a cited paper in-line where you’re reading, including TLDR summaries
Table of Contents to quickly navigate between sections (availability varies)
Save to Library to conveniently track your reading list

Work to expand coverage to more paper sources and add additional features addressing observed challenges is currently in progress

Paper Examples

Here are examples of Semantic Reader operating over popular Computer Science papers across various subfields. We are incrementally improving, testing, and rolling out new features in Semantic Reader so stay tuned. The current design is best experienced on a full-size screen.

NLP
Natural Language Processing

Deep Speech 2: End-to-End Speech Recognition in English and Mandarin

ALBERT: A Lite BERT for Self-supervised Learning of Language Representations

Google’s Multilingual Neural Machine Translation System: Enabling Zero-Shot Translation

CV
Computer Vision

Long-term Recurrent Convolutional Networks for Visual Recognition and Description

Deep Laplacian Pyramid Networks for Fast and Accurate Super-Resolution

Rethinking the Inception Architecture for Computer Vision

ML
Machine Learning

Conditional Generative Adversarial Nets

Learning Important Features Through Propagating Activation Differences

WaveNet: A Generative Model for Raw Audio

Pixel Recurrent Neural Networks

Send us your Semantic Reader feedback.

Semantic Reader is based on research from the Semantic Scholar team at AI2, UC Berkeley and the University of Washington, and supported in part by the Alfred P. Sloan Foundation.

Scim: Intelligent Faceted Highlights for Interactive, Multi-Pass Skimming of Scientific Papers

Raymond Fok, Andrew Head, Jonathan Bragg, Kyle Lo, Marti A. Hearst, Daniel S. Weld

ArXiv

May 9, 2022

Math Augmentation: How Authors Enhance the Readability of Formulas using Novel Visual Design Practices

Andrew Head, Amber Xie, Marti A. Hearst

CHI Conference on Human Factors in Computing Systems

April 29, 2022

From Who You Know to What You Read: Augmenting Scientific Recommendations with Implicit Social Networks

Hyeonsu Kang, Rafal Kocielnik, Andrew Head, Jiangjiang Yang, Matt Latzke, A. Kittur, Daniel S. Weld, Doug Downey, Jonathan Bragg

CHI Conference on Human Factors in Computing Systems

April 21, 2022

CiteRead: Integrating Localized Citation Contexts into Scientific Paper Reading

Napol Rachatasumrit, Jonathan Bragg, Amy X. Zhang, Daniel S. Weld

27th International Conference on Intelligent User Interfaces

March 22, 2022

Paper Plain: Making medical research papers approachable to healthcare consumers with natural language processing

Tal August, L. Wang, Jonathan Bragg, Marti A. Hearst, Andrew Head, Kyle Lo

Preprint

February 28, 2022

TLDR

To improve access to medical papers, we introduce a novel interactive interface-Paper Plain-with four features powered by natural language processing: definitions of unfamiliar terms, in-situ plain language section summaries, a collection of key questions that guide readers to answering passages, and plain language summaries of the answering passages.

SciA11y: Converting Scientific Papers to Accessible HTML

Lucy Lu Wang, Isabel Cachola, Jonathan Bragg, Evie (Yu-Yen) Cheng, Chelsea Hess Haupt, Matt Latzke, Bailey Kuehl, Madeleine van Zuylen, Linda M. Wagner, Daniel S. Weld

ASSETS Demo

October 17, 2021

TLDR

We present SciA11y, a system that renders inaccessible scientific paper PDFs into HTML.

Document-Level Definition Detection in Scholarly Documents: Existing Models, Error Analyses, and Future Directions

Dongyeop Kang, Andrew Head, Risham Sidhu, Kyle Lo, Daniel S. Weld, Marti A. Hearst

EMNLP; Scholarly Document Processing (SDP) Workshop

October 11, 2020

TLDR

The task of definition detection is important for scholarly papers, because papers often make use of technical terminology that may be unfamiliar to readers. We develop a new definition detection system, HEDDEx, that utilizes syntactic features, transformer encoders, and heuristic filters, and evaluate it on a standard sentence-level benchmark.

Augmenting Scientific Papers with Just-in-Time, Position-Sensitive Definitions of Terms and Symbols

Andrew Head, Kyle Lo, Dongyeop Kang, Raymond Fok, Sam Skjonsberg, Daniel S. Weld, Marti A. Hearst

CHI

September 29, 2020

TLDR

We introduce ScholarPhi, an augmented reading interface that brings definitions of technical terms and symbols to readers when and where they need them most.

Experience a smarter way to search and discover scholarly research.

Create Your Account