CVLA assigns CEFR levels to the words based on
CEFR-J Wordlist (Created by Y. Tono). The estimated level of the input text is displayed based on 4 textual features (ARI, VperSent, AvrDiff, and BperA; a detailed explanation will appear after submission). This website is created by
Satoru UCHIDA. This is the second edition of the CVLA, which works faster and shows results per sentence. The results may be slightly different from the former version due to the implementation difference (perl vs. python). Note that the Listening mode is based on scripts of monologues only. The difficulty level varies depending on the quality of the audio, noise, speaking speed, etc. Your feedback is truly appreciated. (kyudai.uchida.lab[at]gmail.com)
[Notice]
CVLA3 Beta is now available. If you are interested, please try it. It would be appreciated if you could provide your feedback via
the form.
CVLA version 1.1 (older version) will no longer be available after August 31st, 2023. If your are redirected to this page, please save this URL for version 2.0.
[Papers]
Uchida, S. and M. Negishi (2018) Assigning CEFR-J levels to English texts based on textual features. In Y. Tono and H. Isahara (eds.) Proceedings of the 4th Asia Pacific Corpus Linguistics Conference (APCLC 2018), pp. 463-467. [
PDF]
内田諭・根岸雅史(2021)「英語読解教材のCEFRレベルの推定 : CVLAの妥当性評価」
Journal of Corpus-based Lexicology Studies, 3, pp.1-14. [
Link]
[History]
Jan, 15, 2023: Due to the server relocation, the security level has been increased. Please note that
special symbols (parentheses (), &, []etc.) cannot be entered.
June, 15, 2021: Fixed the error when the input text has a URL. For security reasons, it is replaced as "replaced-dns", which is counted as a noun.
January 24, 2021: Released.
[Links]
CEFR-J Project
Input text