BioChemQA: A Challenge Dataset for Finding Answers in Biological Publications
As number of scientific papers and specializations grows, one finds it harder to search for facts in publications. Automation of information retrieval may enormously help here and much hope is associated with recent success of deep learning, where we already see some progress in applications for question answering. However, to further advance the state of the art in finding answers from scientific papers we need appropriate challenge datasets. In this ongoing project, we provide a challenge dataset BioChemQA for question answering task, where answers are to be found in Biochemistry related publications.
to be presented at SCIDOCA2017, more description is on the way