Abstrakt
This paper presents a method of improving dependency parsing accuracy by combining four parsers using error-rates (accuracy rates). We use four parsers: MSTParser, MaltParser, TurboParser and MateParser, and the data of the analytical layer of Prague Dependency Treebank.
We parse data with each of the parsers and calculate error-rates for several parameters such as POS of the dependent token. These error-rates are then used to determine weights of edges in an oriented graph created by merging all the parses of a sentence provided by the parsers.
We find a maximum spanning tree in this graph (a dependency tree without cycles), and achieve this way a 1.3 % UAS / 1.1 % LAS improvement compared to the best parser in our experiment.