2000 character limit reached
GU IRLAB at SemEval-2018 Task 7: Tree-LSTMs for Scientific Relation Classification (1804.05408v1)
Published 15 Apr 2018 in cs.CL
Abstract: SemEval 2018 Task 7 focuses on relation ex- traction and classification in scientific literature. In this work, we present our tree-based LSTM network for this shared task. Our approach placed 9th (of 28) for subtask 1.1 (relation classification), and 5th (of 20) for subtask 1.2 (relation classification with noisy entities). We also provide an ablation study of features included as input to the network.
Sponsor
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.
Top Community Prompts
Collections
Sign up for free to add this paper to one or more collections.