Technical notes: Syntax-aware Representation Learning With Pointer Networks
This work-in-progress addresses dependency parsing for natural language processing, but it is incremental as it builds on existing methods like Pointer Networks.
The paper tackles dependency parsing by proposing a novel sequence-to-sequence schema using a BiLSTM and two Pointer Networks with logistic regression, achieving a UAS of 93.14% on the English Penn-treebank, which is 2-3% under state-of-the-art but serves as an attractive baseline.
This is a work-in-progress report, which aims to share preliminary results of a novel sequence-to-sequence schema for dependency parsing that relies on a combination of a BiLSTM and two Pointer Networks (Vinyals et al., 2015), in which the final softmax function has been replaced with the logistic regression. The two pointer networks co-operate to develop a latent syntactic knowledge, by learning the lexical properties of "selection" and the lexical properties of "selectability", respectively. At the moment and without fine-tuning, the parser implementation gets a UAS of 93.14% on the English Penn-treebank (Marcus et al., 1993) annotated with Stanford Dependencies: 2-3% under the SOTA but yet attractive as a baseline of the approach.