Automated interaction between agents necessitates the ability of these agents to discover and select from a set of similar (or identical) services. As a result, trust is used to assess the quality of different cognitive web services. Therefore, this paper proposes artificial intelligence-based English vocabulary test research (AI-EVTR) to overcome the student's requirement. Further, pre-test behavior analysis has been introduced to enhance the apps. Before and after the test period, both groups took a pre-test behavior analysis and post-test log analysis “Vocabulary Test in English.” The experimental group used an app-assisted English vocabulary questionnaire to share their points; they were not extremely motivated in the app-assisted approach using machine learning. Statistical approaches comprising independent samples were used to analyze the acquired data. The experimental group greatly improved between the pre-test and post-test in spelling. Language learning on the website can be an option.