AI-Driven Bilingual Talkbot for Academic Counselling

• Published in: 2023 3rd International Conference on Advance Computing and Innovative Technologies in Engineering (ICACITE) pp. 1986 - 1992
• Main Authors: Krishnam, Nagendra Prasad, Bora, Ashim, Swathi, R.S.V.Rama, Gehlot, Anita, Chandraprakash, V., Raghu, T.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 12.05.2023
• Subjects: academic counseling; bilingual English Arabic; Chatbots; conversational agent; deep learning; NLP; Prediction algorithms; Pupils; Schedules; Software; Talkbot; Terminology; Training
• DOI: 10.1109/ICACITE57410.2023.10182802

Traditional technologies are transforming how businesses interact with their customers, creating demands for prompt responses & continuous accessibility. Students frequently have questions about college & university policies & procedures, academic process, extracurricular, as well as other aspects of educational career. Actually, there is a lack of service satisfaction because the educational advisers as well as the student affairs staff are overloaded with inquiries & unable to respond to them right away. In our work, we develop a multilingual talkbot that converses among learners in both English & Arabic using Artificial Intelligence (AI) & Natural Language Processing (NLP) technologies. The dialog bot was created in Python & is intended for use by students to answer questions about advice. We utilizing a intent built domain-specific compilation as the knowledge source for the talkbots, which is made up of the typical queries students ask advisers & their answers. By analysing the inputs, the talktbot engine ascertains the client's needs and, with an efficiency of 80.00 % in English and 75.00 % in Arabic, finds the most reasonable answer that meets the intention. We also tested the software solution as well as the talkbot's precision using field tests with pupils to see how well it responded to live input.