Abstract
In this paper we present a first-ever manually-built Chinese sentence compression corpus. Based on this corpus, we develop a Chinese sentence compression system and study various measures for evaluation of Chinese sentence compression. We find that 1) using multi-references is very helpful for automatic evaluation in Chinese sentence compression; and 2) besides relational F1, some machine translation evaluation measures are correlated well with human judgments and thus are very promising for future use in this task. © Springer-Verlag 2013.
Author supplied keywords
Cite
CITATION STYLE
Zhang, C., Hu, M., Xiao, T., Jiang, X., Shi, L., & Zhu, J. (2013). Chinese sentence compression: Corpus and evaluation. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8202 LNAI, pp. 257–267). https://doi.org/10.1007/978-3-642-41491-6_24
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
