A survey of syntactic-semantic parsing based on constituent and dependency structures
- PDF / 351,938 Bytes
- 23 Pages / 612 x 792 pts (letter) Page_size
- 104 Downloads / 189 Views
Print-CrossMark
October 2020 Vol. 63 No. 10: 1898–1920 https://doi.org/10.1007/s11431-020-1666-4
Special Topic: Natural Language Processing Technology
. Review .
A survey of syntactic-semantic parsing based on constituent and dependency structures ZHANG MeiShan* School of New Media and Communication, Tianjin University, Tianjin 300072, China Received March 9, 2020; accepted June 3, 2020; published online September 16, 2020
Syntactic and semantic parsing has been investigated for decades, which is one primary topic in the natural language processing community. This article aims for a brief survey on this topic. The parsing community includes many tasks, which are difficult to be covered fully. Here we focus on two of the most popular formalizations of parsing: constituent parsing and dependency parsing. Constituent parsing is majorly targeted to syntactic analysis, and dependency parsing can handle both syntactic and semantic analysis. This article briefly reviews the representative models of constituent parsing and dependency parsing, and also dependency graph parsing with rich semantics. Besides, we also review the closely-related topics such as cross-domain, cross-lingual and joint parsing models, parser application as well as corpus development of parsing in the article. syntax parsing, semantic parsing, constituent parsing, dependency parsing, semantic graph parsing Citation:
Zhang M S. A survey of syntactic-semantic parsing based on constituent and dependency structures. Sci China Tech Sci, 2020, 63: 1898–1920, https://doi.org/10.1007/s11431-020-1666-4
1 Introduction Sentence-level syntactic and semantic parsing is one major topic in the natural language processing (NLP) community, which aims to uncover the internal structural relations in sentences [1–4]. On the one hand, from the view of linguistics, the goal of parsing is to disclose how words are combined to form sentences and the rules that govern the formation of sentences. On the other hand, from the view of NLP applications, parsing can be beneficial for a number of tasks, such as machine translation, question answering, information extraction, sentiment analysis and generation [5–7], and the performance of parsing matters greatly. Parsing has been extensively studied for decades. The goal of syntactic parsing is to derive the syntax information in sentences, such as the subjects, objects, modifiers and topics. There have been a number of achievements for the task, *Corresponding author (email: [email protected])
and large-scale corpora for a range of languages have been already available. Compared with syntactic parsing, semantic parsing is much more difficult due to the complex structure of various semantics such as predicate-argument, and it is also a long-range goal of NLP. With the recent advance in data-driven machine learning models, semantic parsing has received increasing interests, especially under the neural setting. Several datasets based on certain formalizations have been developed to facilitate research. Parsing often relies on spe
Data Loading...
