Interpreting 100,000 Climate Impact Studies with NLP & Generative QA with LFQA

It's time for our next virtual meetup on June 23rd at 7pm. 🎉 Two talks are scheduled: "Interpreting 100,000 Climate Impact Studies with NLP" by Max Callaghan and "Generative Question Answering with LFQA" by Vladimir Blagojevic. As always, we will have smaller breakout room sessions after the talks to connect and discuss. Save the date! Looking forward to meeting you!

"Interpreting 100,000 Climate Impact Studies with NLP"
by Max Callaghan, postdoc at the Mercator Research Institute on Global Commons and Climate Change
Max conducts research on quantitative methods to interpret large collections of scientific literature on climate change. At the intersection of NLP and climate science, his talk describes how language models can be used to classify studies on observed climate impacts. Together with his co-authors, he has created a map of evidence on and attribution of these impacts, providing the most comprehensive picture of the scientific literature on climate change to date.
Paper preprint: https://www.researchsquare.com/article/rs-783398/v1
Interactive maps: https://mcallaghan.github.io/interactive-impacts-map/
Code: https://github.com/mcallaghan/regional-impacts-map

"Generative Question Answering with LFQA"
by Vladimir Blagojevic, long-time open-source contributor and recently joined Senior Software Engineer at deepset
Long-Form Question Answering (LFQA) is a form of generative question answering, which Vladimir implemented in Haystack based on the original research paper. In his talk, he will explain how LFQA systems can first retrieve documents relevant to a user's query and then synthesize accurate, multi-sentence answers. We will be able to try this out with the demo system "Wikipedia Assistant", which is publicly available along with the code and data that was used to train the underlying model. Vladimir will end his talk with insights into the future of generative QA gained at the recent ACL research conference.
Tutorial: https://haystack.deepset.ai/tutorials/lfqa
Original Paper: https://aclanthology.org/P19-1346/