1 Introduction

Introduced the definition, classification, and application of corpora in academic research. Universal corpora aim to comprehensively represent a language, while specialized corpora target specific domains. The use of legal academic English corpora is not yet common in China, despite the existence of legal literature databases and parallel Chinese English corpora. The existing research on academic English in law is mostly based on small-scale samples and lacks large-scale specialized legal corpora. The construction of the LawDEAP Academic English Corpus aims to promote communication between Chinese scholars and international legal research. Its design and construction follow the overall framework of the DEAP Academic English Corpus and can be used for cross disciplinary comparative research.

2. Database construction objectives

LawDEAP aims to build a large-scale and highly timely academic English corpus that comprehensively covers 11 secondary disciplines of law, exploring the principles of database construction and corpus processing methods. This corpus aims to support the teaching and research of legal academic English, providing authentic language materials for vocabulary compilation, language research, and textbook development. The total storage capacity is 5 million words, which actually contains 5014908 words and consists of 697 texts, mainly including research papers and review papers, with 10 sub libraries.

3 Corpus collection

In the process of creating the LawDEAP Law Academic English Corpus, the corpus collection follows four standards: specificity, authority, representativeness, and timeliness. Specialization is reflected in selecting international journals corresponding to the second level discipline of law based on the Chinese degree granting and professional talent training catalog. Authority is ensured by selecting legal journal articles from the Cambridge Academic Journals and ScienceDirect databases, and combining legal expert opinions. In terms of representativeness, research papers and review papers are mainly collected, with research papers accounting for 87.66%. In terms of timeliness, the corpus includes articles published between 2017 and 2019 to reflect the synchronic characteristics of legal academic language. The specific workflow includes downloading articles from a designated database, matching secondary disciplines and journals, reviewing the matching results by a law doctoral teacher, randomly sampling 697 paper corpora, and categorizing them for storage.

4 Text Naming and Meta Information Registration

Introduced the text naming rules and metadata registration methods in the LawDEAP Law Academic English Corpus. The text naming follows the format of "first level discipline+second level discipline+literature number+article type", where the first level discipline is represented by a capital letter L, the second level discipline is abbreviated by two capital letters, and the literature number is three digits. The article types are divided into research papers (RA) and review papers (RV). The textual metadata includes file name, sub library name, journal name, paper title, category, publication year, and DOI number, which are manually summarized and saved as Excel files.

5 Format Conversion and Text Cleaning

Introduced the format conversion and text cleaning work in the process of creating a legal academic English corpus. Firstly, it is necessary to convert the legal journal articles in PDF format to UTF-8 encoded text format for easy retrieval and data extraction. Use tools such as Adobe Acrobat Pro and ABBYY FineReader 14 for format conversion, and manually clean up the converted text, including deleting irrelevant information, correcting errors, and garbled text. The text cleaning work was jointly completed by five staff members and divided into three rounds to ensure the quality and consistency of the corpus. Finally, use PowerGREP software to clear the footnote embedding issue and count the total amount of corpus to complete the storage.

6. The Later Development and Application Prospects of LawDEAP

The post development of LawDEAP Law Academic English Corpus aims to expand the corpus size and increase the depth of corpus processing. The plan is to achieve this by collecting more legal journal corpora from databases, including academic monographs and conference papers, and expanding the time span of corpus collection. At the same time, corpus processing will include various deep processing methods such as part of speech tagging and syntactic structure tagging. LawDEAP has broad application prospects and can be used to generate index rows, word lists, calculate collocations, topic words, etc., serving legal researchers, language teachers, and legal English learners. It supports academic English research in law, including analyzing academic hotspots, rhetorical features, lexical and syntactic features, as well as language differences in different disciplinary fields. In addition, LawDEAP can be used to create academic word lists or terminologies for legal secondary disciplines, compare and analyze the language characteristics of each discipline, and summarize the whole sentence definitions and high-frequency collocation forms of vocabulary or phrases based on the perspective of genre and phrase studies, providing support for the compilation of dictionaries for legal academic English learners.

7 Conclusion

LawDEAP, as a sub library of DEAP, follows the standards of the main library and selects corpus from authoritative journals and legal research articles based on the opinions of legal experts. The project team has clear division of labor to ensure the representativeness, authenticity, and accuracy of the corpus. The corpus covers a wide range of legal fields, balancing scale and timeliness, reflecting the characteristics of legal academic English. Provide valuable materials in legal language research, teaching, and dictionary compilation. The corpus is open and supports researchers to further process and update it.