Building a Korean Web Corpus for Analyzing Learner Language
Markus Dickinson, Ross Israel, Sun-Hee Lee · 2010
Post-positional particles are a significant source of errors for learners of Korean. Following methodology that has proven effective in handling English preposition errors, we are beginning the process of building a machine learner for particle error detection in L2 Korean writing. As a first step, however, we must acquire data, and thus we present a methodology for constructing large-scale corpora of Korean from the Web, exploring the feasibility of building corpora appropriate for a given topic and grammatical construction. 1