We introduce Talk to Papers, which exploits the recent open-domain question answering (QA) techniques to improve the current experience of academic search. It's designed to enable researchers to use natural language queries to find precise answers and extract insights from a massive amount of academic papers. We present a large improvement over classic search engine baseline on several standard QA datasets and provide the community a collaborative data collection tool to curate the first natural language processing research QA dataset via a community effort.

Advances in AI are changing the way we search, socialize and work online. How would research change if we applied these advances to academic search? This is the question that drives our work on Semantic Scholar, a free non-profit academic search engine with a mission to help scientists fight information overload. And as part of our commitment to contributing to the common good through technology, we want to start sharing our experiences more broadly. As a technical team at an AI-focused nonprofit, we have the expertise of a large company, the work ethic of a startup, and the challenges of a nonprofit.