1 Introduction

The itGLOBE corpus is part of the BFSU Global Corpus Cluster project, which includes 101 languages and uses the Brown corpus sampling scheme to promote bilingual or multilingual comparative research. This article introduces the background, process, and application value of the itGLOBE corpus.

2. Construction process of Italian language corpus

The construction of the Italian language corpus began in the late 1940s, initiated by Roberto Bussa, whose computer indexing of the Latin works of Saint Thomas Aquinas marked the birth of the first machine-readable corpus. The development of the Italian language corpus has gone through three stages: the initial stage (early 1950s to early 1970s), the dictionary compilation stage (early 1970s to late 1990s), and the comprehensive development stage (late 1990s to present). In the initial stage, the goal is to compile a work index, while in the dictionary compilation stage, dictionaries are compiled based on corpora. In the comprehensive development stage, a diversified and functionally rich corpus pattern is formed. The Italian language corpus plays an important role in the classification, storage, and protection of language resources, and is widely used in language learning, research, and resource record protection. Natural language processing technology is reflected in corpus construction, and there is continuity among various studies. The creation of the itGLOBE corpus aims to fill the gap in direct comparison between Italian, English, and Chinese corpora, and provide balanced corpus resources for Italian language research.

Creation of 3 itGLOBE corpora

The itGLOBE corpus is a million word balanced corpus that collects written Italian language versions published after 2013. It has been uploaded to the CQPweb multilingual corpus platform of BFSU and provides various online analysis functions. The corpus sampling scheme refers to the Brown Corpus, covering four genre types: news, general, academic, and novel, as well as 15 sub genres, with a particular focus on collecting local detective novels. Corpus collection includes text selection, input, and meta information annotation. The texts are all original Italian texts, excluding translated texts, with a wide time range covering texts after 2000. The corpus provides three versions: raw corpus, part of speech encoding, and word form restoration, to meet different research needs. All corpus is manually proofread, stored in UTF-8 encoding format, and accompanied by detailed metadata. The platform supports simple queries and CQP syntax to meet the needs of different users.

4 Research Prospects

The balance and representativeness of the itGLOBE corpus make it an important resource for studying the characteristics of Italian language after 2000. The word frequency table and n-gram table are helpful for course design and dictionary compilation. The four major categories of texts, namely news, general, academic, and fiction, can support research on discourse genre and domain variation. The comparability of itGLOBE enables it to be compared with other corpora, such as conducting Italian English comparisons with CROWN2021, Italian Chinese comparisons with ToRCH series, and conducting typological studies with other GLOBE corpora. The sampling scheme of itGLOBE is similar to the Italian reference library, which is helpful for studying changes over time. The novel sub library can be used for translation studies in parallel with the Italian Chinese corpus. The subsequent corpus construction includes updating corpus metadata, reducing the proportion of corpus before 2013, expanding to support more comparative research, and considering adding oral corpus and emerging genres.

5 Conclusion

The development of Italian language corpus is mature, and balanced corpus is valuable for multilingual comparative research; The itGLOBE corpus adopts the Brown corpus sampling method, which is compatible with the Brown family and GLOBE corpus; Can be used online on the CQPweb platform of Beijing Foreign Studies University to meet the needs of different users; Following the concept of co construction and sharing, encouraging peer participation, and promoting Italian language teaching and research.