Is There a Core General Vocabulary? Introducing the "New General Service List".
By V. Brezina, Dana Gablasova. 2015
306 citations recorded by the source corpus.
The current study presents a New General Service List (new-GSL), which is a result of robust comparison of four language corpora (LOB, BNC, BE06, and EnTenTen12) of the total size of over 12 billion running words. The four corpora were selected to represent a variety of corpus sizes and approaches to representativeness and sampling. In particular, the study investigates the lexical overlap among the corpora in the top 3,000 words based on the average reduced frequency (ARF), which is a measure that takes into consideration both frequency and dispersion of lexical items. The results show that there exists a stable vocabulary core of 2,122 items (70.7%) among the four corpora. Moreover, these vocabulary items occur with comparable ranks in the individual wordlists. In producing the new-GSL, the core vocabulary items were combined with new items frequently occurring in the corpora representing current language use (BE06 and EnTenTen12). The final product of the study, the new-GSL, consists of 2,494 lemmas and covers between 80.1 and 81.7 per cent of the text in the source corpora.
Subfields: Artificial Intelligence, Developmental and Educational Psychology
View Paper: https://www.semanticscholar.org/paper/2a8ea9e0131ff026c288583e5da724672a071fef
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
In Developmental and Educational Psychology
Citations in the corpus
- Advancing disciplinary literacy through English for academic purposes: Discipline-specific wordlists, collocations and word families for eight secondary subjects (2018, methodology): “However, despite their different methods for operationalizing the construct of academic vocabulary and not using the original GSL, we also computed the matches between the AVL and New GSL and found 493 shared lemmas, i.e. approximately 16% of this academic wordlist overlaps with a wordlist represent”