The aim of relation classification is to categorize into predefined classes the relations btw pairs of given entities in texts.
There are two ways to represent relations between entities using neural networks: recurrent neural networks (RNNs) and convolutional neural networks (CNNs), We tried RNNs in Tensorflow.
We implemented three LSTM-RNN architectures:
- Classify using LSTM Networks along Shortest Dependency Paths.
- Classify using bidirectional LSTM Networks on LCA Sub Tree.
- Classification using LSTMS on Sequences and Tree Structures.
Better performance is achived in Classification using LSTMS on Sequences and Tree Structures..