pith. machine review for the scientific record. sign in

arxiv: 1504.06580 · v2 · submitted 2015-04-24 · 💻 cs.CL · cs.LG· cs.NE

Recognition: unknown

Classifying Relations by Ranking with Convolutional Neural Networks

Authors on Pith no claims yet
classification 💻 cs.CL cs.LGcs.NE
keywords taskclassificationfeaturesrankingachieveartificialclassifyingconvolutional
0
0 comments X
read the original abstract

Relation classification is an important semantic processing task for which state-ofthe-art systems still rely on costly handcrafted features. In this work we tackle the relation classification task using a convolutional neural network that performs classification by ranking (CR-CNN). We propose a new pairwise ranking loss function that makes it easy to reduce the impact of artificial classes. We perform experiments using the the SemEval-2010 Task 8 dataset, which is designed for the task of classifying the relationship between two nominals marked in a sentence. Using CRCNN, we outperform the state-of-the-art for this dataset and achieve a F1 of 84.1 without using any costly handcrafted features. Additionally, our experimental results show that: (1) our approach is more effective than CNN followed by a softmax classifier; (2) omitting the representation of the artificial class Other improves both precision and recall; and (3) using only word embeddings as input features is enough to achieve state-of-the-art results if we consider only the text between the two target nominals.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.