Numerical reasoning is a complex subtask in machine reading comprehension, which requires the model to complete discrete operations such as statistics and calculation in the process of reasoning answers. As far as we know, the best solution for numerical reasoning adopts GNN structure. However, the graph structure is often generated by hand-designed rules and it will cost extra computation resource. To tackle these problems, we propose a Transformer-based numerical reasoning solution which transforms the hand-designed graph structure into learnable parameters. Extensive experimental results show that compared to those GNN-based methods, our method can carry out numerical reasoning more efficiently only at the cost of a little decrease on answer quality.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.