Computational linguistics: a scientometric review

Computational linguistics has significantly influenced teaching, learning, and research methods in knowledge production and data analysis. This study aims to examine the rise and development of computational linguistics and its implications for education. Using a scientometric approach, we analysed 81,740 documents from Scopus, Web of Science (WOS), and Lens between 1954 and 2022. Bibliometric and scientometric analyses were conducted using CiteSpace 5.8.R3 and VOSviewer 1.6.18. Our analysis identified eight bibliometric and eight scientometric indicators that describe the development of computational linguistics. We observed two primary patterns in the clustering of the most explored and examined topics within computational linguistics. The first pattern includes using computational linguistics to produce progress reports, advance linguistic theory, and assess inter-coder agreement. The second pattern involves utilizing probabilistic language models to evaluate word sense disambiguation in semantic relatedness, deep learning, semantic roles, referring expressions, and statistical machine translation. Our analysis reveals: (1) The development of natural language processing applications to enhance learning experiences and analyse student interactions; (2) The creation of intelligent tutoring systems and chatbots for personalized guidance and support; (3) The implementation of automated assessment and grading systems for consistency and timely feedback; (4) The role of computational linguistics in language learning tools and resources, providing real-time feedback and adaptive learning paths; and (5) The use of computational linguistics methods to analyse research trends and emerging topics. By understanding the development and contributions of computational linguistics, we can better harness its potential to enhance the quality of education, benefiting students, educators, and researchers alike.