next up previous
Next: Parsing Up: The Alpino Dependency Treebank Previous: Dependency Trees

The annotation process

The annotation process is roughly divided into two parts: we first parse a sentence with the Alpino parser and then select the best parse from the set of generated parses. Several tools that we have developed and implemented in Hdrug, a graphical environment for natural language processing (van Noord and Bouma 1997), facilitate the two parts of the annotation process. In section 4.1 we present an interactive lexical analyzer, a constituent marker and a tool for temporary addition of lexical information. The parse selection tool is described in in section 4.2.


Noord G.J.M. van