OpenIE6: Iterative Grid Labeling and Coordination Analysis for Open Information Extraction

Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, pages 3748–3761, November 16–20, 2020. c⃝2020 Association for Computational Linguistics 3748 OpenIE6: Iterative Grid Labeling and Coordination Analysis for Open Information Extraction Keshav Kolluru1*, Vaibhav Adlakha1*, Samarth Aggarwal1, Mausam1, and Soumen Chakrabarti2 1 Indian Institute of Technology Delhi keshav.kolluru@gmail.com, vaibhavadlakha95@gmail.com samarth.aggarwal.2510@gmail.com, mausam@cse.iitd.ac.in 2 Indian Institute of Technology Bombay soumen@cse.iitb.ac.in Abstract A recent state-of-the-art neural open informa- tion extraction (OpenIE) system generates ex- tractions iteratively, requiring repeated encod- ing of partial outputs. This comes at a signif- icant computational cost. On the other hand, sequence labeling approaches for OpenIE are much faster, but worse in extraction quality. In this paper, we bridge this trade-off by pre- senting an iterative labeling-based system that establishes a new state of the art for OpenIE, while extracting 10× faster. This is achieved through a novel Iterative Grid Labeling (IGL) architecture, which treats OpenIE as a 2-D grid labeling task. We improve its performance fur- ther by applying coverage (soft) constraints on the grid at training time. Moreover, on observing that the best OpenIE systems falter at handling coordination struc- tures, our OpenIE system also incorporates a new coordination analyzer built with the same IGL architecture. This IGL based coordina- tion analyzer helps our OpenIE system han- dle complicated coordination structures, while also establishing a new state of the art on the task of coordination analysis, with a 12.3 pts improvement in F1 over previous analyzers. Our OpenIE system, OpenIE61, beats the pre- vious systems by as much as 4 pts in F1, while being much faster. 1 Introduction Open Information Extraction (OpenIE) is an ontology-free information extraction paradigm that generates extractions of the form (subject; rela- tion; object). Built on the principles of domain- independence and scalability (Mausam, 2016), OpenIE systems extract open relations and argu- ments from the sentence, which allow them to be *Equal Contribution 1https://github.com/dair-iitd/openie6 used for a wide variety of downstream tasks like Question Answering (Yan et al., 2018; Khot et al., 2017), Event Schema Induction (Balasubramanian et al., 2013) and Fact Salience (Ponza et al., 2018). Figure 1: The extractions (Rome; [is] the capital of; Italy) and (Rome; is known for; it’s rich history) can be seen as the output of grid labeling. We additionally introduce a token [is] to the input. End-to-end neural systems for OpenIE have been found to be more accurate compared to their non- neural counterparts, which were built on manually defined rules over linguistic pipelines. The two most popular neural OpenIE paradigms are gener- ation (Cui et al., 2018; Kolluru et al., 2020) and labeling (Stanovsky et al., 2018; Roy et al., 2019). Generation systems generate extractions one word at a time. IMoJIE (Kolluru et al., 2020) is a state-of-the-art OpenIE system that re-encodes the partial set of extractions output thus far when generating the next extraction. This captures de- pendencies among extractions, reducing the overall redundancy of the output set. However, this re- peated re-encoding causes a significant reduction in speed, which …

Content truncated to 500 words