Paper

Combining Neural Networks and Log-linear Models to Improve Relation Extraction

The last decade has witnessed the success of the traditional feature-based method on exploiting the discrete structures such as words or lexical patterns to extract relations from text. Recently, convolutional and recurrent neural networks has provided very effective mechanisms to capture the hidden structures within sentences via continuous representations, thereby significantly advancing the performance of relation extraction... (read more)

Results in Papers With Code
(↓ scroll down to see all results)