Emergent communication (EC) is the field that seeks to understand the mechanisms behind the emergence and evolution of natural language. In EC, the de facto standard has been using sequential architectures that have not explicitly incorporated the "tree-structured hierarchy" inherent in human language. This study utilizes a stack-based model called RL-SPINN, which learns tree structures through reinforcement learning without ground-truth parsing data, and acquires sentence representations according to these structures. We use this model as the basis for the understanding agents and investigate the extent to which the inductive bias of an architecture that explicitly utilizes tree structures affects the emergent language. The experimental results show that the emergent language generated by our model exhibits higher communication accuracy than those generated by other baselines in some settings. This work is the first to focus on the tree-structured hierarchy of language and suggests new directions for future research in EC.

Motivation

While text mining technologies for biomedical research have gained popularity as a way to take advantage of the explosive growth of information in text form in biomedical papers, selecting appropriate natural language processing (NLP) tools is still difficult for researchers who are not familiar with recent advances in NLP. This article provides a comparative evaluation of several state-of-the-art natural language parsers, focusing on the task of extracting protein-protein interaction (PPI) from biomedical papers. We measure how each parser, and its output representation, contributes to accuracy improvement when the parser is used as a component in a PPI system.

Results

All the parsers attained improvements in accuracy of PPI extraction. The levels of accuracy obtained with these different parsers vary slightly, while differences in parsing speed are larger. The best accuracy in this work was obtained when we combined Miyao and Tsujii's Enju parser and Charniak and Johnson's reranking parser, and the accuracy is better than the state-of-the-art results on the same data.

Availability

The PPI extraction system used in this work (AkanePPI) is available online at http://www-tsujii.is.s.u-tokyo.ac.jp/downloads/downloads.cgi. The evaluated parsers are also available online from each developer's site.