Erik Tjong Kim Sang, email@example.com
Text chunking consists of finding non-recursive non-overlapping phrases. We will present a method for detecting arbitrary noun phrases by performing a repeated noun phrase chunking process on the data. A memory-based program is used for recognizing base noun phrases and using these for building higher order noun phrases. This method has been applied to the standard data set for the CoNLL'99 workshop and has obtained precision and recall rates of 91.3% and 76.1% (crossing rate 0.1).