Application of k-means clustering in spoken English training process

Abstract

Data Clustering is the most important data mining technique playing a vital role in various fields such as Business,
Medicine, Construction, etc. In this study, k-means clustering technique is utilized to understand the skill level of
the students enrolled in Spoken English Training (SET) programme and effectively strategize the need-based
training sessions for them according to their present knowledge and requirements. Pre-training data collected
from 159 students enrolled for Spoken English Training (SET) programmes in Chennai, India which consists of
marks secured by the students in three tests concerning five basic categories namely content, communication,
pronunciation, vocabulary, and grammar. All the necessary skills required in each category are thoroughly
examined and scored. Data is then clustered using Elbow method and clustering technique to categorize the
student mass into four different groups. The strengths and weaknesses of each group are uniquely diagnosed
and necessary tailor-made curriculum and training sessions are advised so that effective suitable training can be
given to each candidate at optimal time duration and cost.