Improving Chinese Semantic Role Labeling using High-quality Surface and Deep Case Frames

Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 1, Long Papers, pages 568–577, Valencia, Spain, April 3-7, 2017. c⃝2017 Association for Computational Linguistics Improving Chinese Semantic Role Labeling using High-quality Surface and Deep Case Frames Gongye Jin∗, Daisuke Kawahara, Sadao Kurohashi Graduate School of Informatics, Kyoto University Yoshida-honmachi, Sakyo-ku, Kyoto, 606-8501, Japan jin@nlp.ist.i.kyoto-u.ac.jp, {dk, kuro}@i.kyoto-u.ac.jp Abstract This paper presents a method for improv- ing semantic role labeling (SRL) using a large amount of automatically acquired knowledge. We acquire two varieties of knowledge, which we call surface case frames and deep case frames. Although the surface case frames are compiled from syntactic parses and can be used as rich syntactic knowledge, they have limited ca- pability for resolving semantic ambiguity. To compensate the deficiency of the sur- face case frames, we compile deep case frames from automatic semantic roles. We also consider quality management for both types of knowledge in order to get rid of the noise brought from the automatic anal- yses. The experimental results show that Chinese SRL can be improved using au- tomatically acquired knowledge and the quality management shows a positive ef- fect on this task. 1 Introduction Semantic role labeling (SRL) is regarded as a task that is intermediate between syntactic analysis and semantic analysis in natural language processing (NLP). The main goal of SRL is to extract a propo- sition from a sentence about who does what to whom, when, where and why. By using semantic roles, the complex expression of a sentence is then interpreted as an event and its participants (i.e., a predicate and arguments such as agent, patient, locative, temporal and manner). Unlike syntactic level surface cases (i.e., dependency labels such as subject and object), semantic roles can be regarded ∗The first author is now affiliated with Canon IT Solu- tions Inc. as a deep case representation for predicates. Be- cause of its ability to abstract the meaning of a sen- tence, SRL has been applied to many NLP appli- cations, including information extraction (Chris- tensen et al., 2010), question answering (Pizzato and Moll´a, 2008) and machine translation (Liu and Gildea, 2010). Semantically annotated corpora, such as FrameNet (Fillmore et al., 2001) and PropBank (Kingsbury and Palmer, 2002), make this type of automatic semantic structure analysis feasible by using supervised machine learning methods. However, supervised SRL methods have the following two major issues. Firstly, as a common issue in almost all the supervised approaches, it is expensive to enlarge manually annotated corpora to learn a more accurate model. Secondly, experiments show that automatic SRL systems strongly depend on syntactic information. In practice, these SRL systems suffer from errors propagated from the lower-level syntactic anal- yses, such as word segmentation, POS tagging, and dependency parsing. Although some studies use automatic analyses of unlabeled corpora to enrich the training data to solve the first problem (F¨urstenau and Lapata, 2009), accumulated errors in such automatic analysis inevitably cause nega- tive effects. Especially, for some hard-to-analyze languages, such …

Content truncated to 500 words