We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CL

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computation and Language

Title: Relation Extraction with Self-determined Graph Convolutional Network

Abstract: Relation Extraction is a way of obtaining the semantic relationship between entities in text. The state-of-the-art methods use linguistic tools to build a graph for the text in which the entities appear and then a Graph Convolutional Network (GCN) is employed to encode the pre-built graphs. Although their performance is promising, the reliance on linguistic tools results in a non end-to-end process. In this work, we propose a novel model, the Self-determined Graph Convolutional Network (SGCN), which determines a weighted graph using a self-attention mechanism, rather using any linguistic tool. Then, the self-determined graph is encoded using a GCN. We test our model on the TACRED dataset and achieve the state-of-the-art result. Our experiments show that SGCN outperforms the traditional GCN, which uses dependency parsing tools to build the graph.
Comments: CIKM-2020
Subjects: Computation and Language (cs.CL)
Cite as: arXiv:2008.00441 [cs.CL]
  (or arXiv:2008.00441v2 [cs.CL] for this version)

Submission history

From: Sunil Sahu [view email]
[v1] Sun, 2 Aug 2020 09:52:58 GMT (7233kb,D)
[v2] Thu, 27 Aug 2020 05:55:43 GMT (161kb,D)

Link back to: arXiv, form interface, contact.