The Resolution of Gender Anaphora Reference with the Help of Kernel Trick Mechanisms

2019

Full text Open access: Gold Metadata (CSL-JSON)
DOI 10.1088/1757-899x/630/1/012029
Data confidence 85 high
How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

Abstract An article represents the results of the experiments constructing of conventional anaphora resolution system for gender-balanced corpora GAP. The main idea behind the conducted set of experiments is taking advantage over the kernel trick mechanism of deep word vectorization for various semantic and syntax features. We will show that it has a positive impact on system performance versus the traditional approach.