Korean Text Chunk Identification Using Support Vector Machines

In this paper, we propose a method of Korean text chunk identification based on support vector machines (SVMs). Text chunking is a task that divides text into syntactically related non-overlapping groups of word. It is a useful preprocessing step for the reduced time and computational resource of sentence parsing. Especially, we select features for SVM by considering the linguistic typological characteristics of Korean, and convert the text chunk identification, a multi-class problem into binary problems for the use of SVM frameworks. According to the experimental results, the proposed method showed 99.04 of F-score