ChatSense – A Multilingual Chatbot
M K Saravana, Arman Mohammed, Sovan Pattanayak, Devendra Kumar Paswan, Yuvraj Dadhich · International Journal of Scientific Research in Science and Technology · 2025
Multilingual chatbots have become essential tools for global communication, enabling seamless interaction across diverse languages and cultures. This paper presents the implementation of "ChatSense," a multilingual AI assistant leveraging pre-trained language models (PLMs) such as Gemini-2.0-flash, Google Translate APIs, and speech recognition technologies. The chatbot supports over 100 languages, addressing challenges like low- resource language support, contextual accuracy, and cultural adaptability. Our implementation integrates PLMs with knowledge graphs and fine-tuning strategies, including adapter-based tuning and prompt engineering, to enhance performance. Experimental results demonstrate significant improvements in response accuracy and user satisfaction across multiple languages, particularly in healthcare and educational domains. Graphical analyses illustrate the chatbot's effectiveness in handling cross-lingual queries and maintaining contextual coherence. This work contributes to advancing inclusive and efficient conversational AI systems.