KatzBot: Revolutionizing Academic Chatbot for Enhanced Communication

TL;DR

KatzBot, an innovative chatbot powered by KatzGPT, a custom Large Language Model (LLM) fine-tuned on domain-specific academic data that outperforms established existing open source LLMs, achieving higher accuracy and domain relevance.

Abstract

Effective communication within universities is crucial for addressing the diverse information needs of students, alumni, and external stakeholders. However, existing chatbot systems often fail to deliver accurate, context-specific responses, resulting in poor user experiences. In this paper, we present KatzBot, an innovative chatbot powered by KatzGPT, a custom Large Language Model (LLM) fine-tuned on domain-specific academic data. KatzGPT is trained on two university-specific datasets: 6,280 sentence-completion pairs and 7,330 question-answer pairs. KatzBot outperforms established existing open source LLMs, achieving higher accuracy and domain relevance. KatzBot offers a user-friendly interface, significantly enhancing user satisfaction in real-world applications. The source code is publicly available at \url{https://github.com/AiAI-99/katzbot}.

Figures (3)

• Figure 1: Train and test dataset embedding space
• Figure 2: Katzbot interface
• Figure 3: The katzGPT model features N Transformer decoder blocks. Each block includes a multi-head masked attention layer, a multi-layer perceptron layer, normalization, dropout layers, and utilizes residual connections to learn from the previous block's input. The multi-head masked attention layer captures sequential relationships in the input sequence using Q, K, and V vectors.