A Computational Model of Korean Mental Lexicon

This paper proposes a computational model of Korean mental lexicon which can explain very basic findings that are observed using the visual Korean lexical decision task. The model is Frequency-based trie which is based on the trie data structure. Alphabets in each node of the Frequency-based trie is sorted in descending order by by their frequency in the corpus. It is the main idea which enables the model to simulate the Korean mental lexicon. There will be evaluations and comparison results with human data: how well the model simulates Korean word processing and how much the simulated results coincide with that of human processing. We also discuss both the strength and weakness of the computational model.