Evaluation of Domain-Specific Vocabulary with Machine Learning-Based Techniques: Japanese and Russian Case Studies

Foreign language education is one of the prominent requirements. Undergraduate students at the Facultyof Tourism are offered the opportunity to learn a second foreign language, which will contribute to theirprofessional lives. However, this second foreign language, which is taught from the beginner level, cannotcontribute to the students’ professional lives at a desired level unless it includes professional technicalterms related to their profession. For this reason, foreign language education books should include fieldwords related to the professional field to a certain extent. This study examines the suitability of foreignlanguage education books used at the basic level in Russian and Japanese courses from the scope oftheir field speciality. First, the frequently used words in the fields of “Tourism and Hotel Management”and “Tourism Guidance” were determined and set as the keywords. Then, depending on these keywords,other frequently used words were obtained using machine learning and natural language processingtechniques. For this purpose, we used Python’s Gensim library, and we established corpuses of wordvectors consisting of both the keywords and the near-distanced words to these keywords in each field withthe help of pre-trained word vector models. This study revealed statistically to what extent the textbookscurrently used contain the domain-specific vocabulary in the field.