TY - JOUR
T1 - Evaluation of feature combination for effective structural disambiguation
AU - Park, So Young
AU - Kwak, Yong Jae
AU - Lim, Joon Ho
AU - Rim, Hae Chang
PY - 2004
Y1 - 2004
N2 - In this paper, we present the useful features of a syntactic constituent for a probabilistic parsing model and analyze the combination of the features in order to disambiguate parse trees effectively. Unlike most of previous works focusing on the features of a single head, the features of a functional head, the features of a content head, and the features of size are utilized in this study. Experimental results show that the combination of different features such as the functional head feature and the size feature is prefered to the combination of similar features such as the functional head feature and the content head feature. Besides, it is remarkable that the function feature is more useful than the combination of the content feature and the size feature.
AB - In this paper, we present the useful features of a syntactic constituent for a probabilistic parsing model and analyze the combination of the features in order to disambiguate parse trees effectively. Unlike most of previous works focusing on the features of a single head, the features of a functional head, the features of a content head, and the features of size are utilized in this study. Experimental results show that the combination of different features such as the functional head feature and the size feature is prefered to the combination of similar features such as the functional head feature and the content head feature. Besides, it is remarkable that the function feature is more useful than the combination of the content feature and the size feature.
UR - http://www.scopus.com/inward/record.url?scp=35048823897&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=35048823897&partnerID=8YFLogxK
M3 - Article
AN - SCOPUS:35048823897
VL - 2945
SP - 97
EP - 101
JO - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
JF - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SN - 0302-9743
ER -