Long Short-Term Memory Over Tree Structures
This addresses a fundamental problem in natural language understanding by enabling more effective semantic composition through tree-structured memory, though it is an incremental extension of existing LSTM methods.
The paper tackles the problem of modeling long-distance interactions in hierarchical structures like language or image parses by extending LSTM to tree structures, resulting in the S-LSTM model that outperforms a state-of-the-art recursive model in semantic composition tasks.
The chain-structured long short-term memory (LSTM) has showed to be effective in a wide range of problems such as speech recognition and machine translation. In this paper, we propose to extend it to tree structures, in which a memory cell can reflect the history memories of multiple child cells or multiple descendant cells in a recursive process. We call the model S-LSTM, which provides a principled way of considering long-distance interaction over hierarchies, e.g., language or image parse structures. We leverage the models for semantic composition to understand the meaning of text, a fundamental problem in natural language understanding, and show that it outperforms a state-of-the-art recursive model by replacing its composition layers with the S-LSTM memory blocks. We also show that utilizing the given structures is helpful in achieving a performance better than that without considering the structures.