In Human Computer Interface (HCI) design has long been emphasized. Recent developments in chatbot technology demonstrate a move away from swipes, scrolls, and clicks and toward natural language interfaces that allow text-based interactions. For years, HCI has placed a lot of emphasis on graphic user interface (GUI) design. But the new developments in chatbot technology demonstrate a move toward natural language interfaces, which allow text-based communication in place of swipes and clicks. The wide range of user expectations and behaviors make instructional chatbots designing a challenging task. At present Context and persona-based chatbots for college students are the subject of little research, which emphasizes the need for novel design strategies. This study employed iteration-based design science research (DSR) technique. Machine learning approaches were used for persona elicitation at first. Persona3D, a data-driven approach to persona construction, was used and finally structural equation modeling was incorporated on technology adoption together with previous machine learning findings. Finally, the Persona3D method was realized through the design and construction of several chatbot prototypes using Persona3D models and path mapping. The analysis of machine learning revealed eight different student groups. It was discovered that the main indicators of chatbot usage were performance expectancy, effort expectancy, and habits. Eight chatbots were then created to assist with tasks such as evaluation, assignment advice, knowledge acquisition, and lab support. Several contributions are made by this work, the main one being the Persona3D model and approach. Additionally, this study is the first to look at how persona moderators affect students’ adoption and usage of chatbots.