CTAP for Chinese:A Linguistic Complexity Feature Automatic Calculation Platform

Proceedings of the 13th Conference on Language Resources and Evaluation (LREC 2022), pages 5525–5538 Marseille, 20-25 June 2022 © European Language Resources Association (ELRA), licensed under CC-BY-NC-4.0 5525 CTAP for Chinese: A linguistic Complexity Feature Automatic Calculation Platform Yue Cui123, Junhui Zhu123, Liner Yang123*, Xuezhi Fang123, Xiaobin Chen4, Yujie Wang5, Erhong Yang13 1National Language Resources Monitoring and Research Center Print Media Language Branch, Beijing Language and Culture University, Beijing 100083, China 2School of Information Science, Beijing Language and Culture University, Beijing 100083, China 3Beijing Advanced Innovation Center for Language Resources, Beijing Language and Culture University, Beijing 100083, China 4Hector-Institut f¨ur Empirische Bildungsforschung, T¨ubingen Universit¨at, 72072 T¨ubingen, Germany 5School of Computer and Information Technology, Beijing Jiaotong University, Beijing 100044, China cyue829@163.com, {nysyzxzjh, lineryang, jasonfang3900, chxiaobin}@gmail.com, yujie.wang@bjtu.edu.cn, yerhong@blcu.edu.cn Abstract The construct of linguistic complexity has been widely used in the research of language learning. Several text analysis tools have been made to automatically analyze linguistic complexity. However, the indexes supported by several existing Chinese text analysis tools are limited and varied due to different research purposes. CTAP is an open-source toolkit for linguistic complexity measurement extraction, which serves all research purposes. Although it was originally developed for English, the Unstructured Information Management (UIMA) framework it used allows the integration of other languages. In this study, we integrated the Chinese component into CTAP, describing the index sets it incorporated and comparing it with three linguistic complexity tools for Chinese. The index set includes 4 levels of 196 linguistic complexity indexes: character level, word level, sentence level, and discourse level. So far, CTAP has implemented automatic calculation of complexity characteristics for four languages, aiming to help linguists without NLP background do their research on language complexity. Keywords: linguistic complexity analyzer, Chinese, text analysis 1. Introduction Linguistic complexity is a multifaceted construct used in a wide range of contexts (Bult and Housen, 2014), manifested in the variety and complexity of production units or grammatical structures (Wolfe-Quintero et al., 1998). It has been used in various studies, such as sec- ond language proficiency and development assessment (Crossley and McNamara, 2014; Bult and Housen, 2014; Kyle, 2016), readability assessment (Vajjala and Meurers, 2012; Feng et al., 2010; Chen and Meurers, 2018), and first language academic writing (Crossley et al., 2011; Weiss and Meurers, 2019). A large num- ber of complexity indexes have been proposed in the above research tasks. However, a common and crucial question these studies have to answer is to what ex- tent the complexity indexes that have existed are valid, which would directly bear upon the validity of the re- search results. The best way to solve this problem is to compare the proposed indexes with a large number of texts of varying difficulty (Lu, 2010), which requires reliable computational tools that can automatically cal- culate linguistic complexity index values. Most of the linguistic complexity work has focused on English (Okinina et al., 2020), so many computa- tional tools can automatically analyze English texts, such as Coh-Metrix, D-Level Analyzer, L2 Syntactic Complexity Analyzer, CTAP, …

Content truncated to 500 words