Please Mind the Root: Decoding Arborescences for Dependency Parsing
This solves a specific technical issue in NLP dependency parsing by ensuring grammatical correctness without computational overhead.
The paper addresses the problem of dependency parsers violating the single-root constraint in dependency trees, showing that state-of-the-art parsers have violation rates up to 24%, especially with smaller training sets. They adapt an algorithm from Gabow and Tarjan (1984) to enforce this constraint without increasing decoding runtime.
The connection between dependency trees and spanning trees is exploited by the NLP community to train and to decode graph-based dependency parsers. However, the NLP literature has missed an important difference between the two structures: only one edge may emanate from the root in a dependency tree. We analyzed the output of state-of-the-art parsers on many languages from the Universal Dependency Treebank: although these parsers are often able to learn that trees which violate the constraint should be assigned lower probabilities, their ability to do so unsurprisingly de-grades as the size of the training set decreases. In fact, the worst constraint-violation rate we observe is 24%. Prior work has proposed an inefficient algorithm to enforce the constraint, which adds a factor of n to the decoding runtime. We adapt an algorithm due to Gabow and Tarjan (1984) to dependency parsing, which satisfies the constraint without compromising the original runtime.