Authors :
Surendra Digumarthi; Sarita Padhi; Sai Raghava; Sreekanth Putsala; Shirish Kumar Gonala; Bharani Kumar Depuru
Volume/Issue :
Volume 9 - 2024, Issue 2 - February
Google Scholar :
https://tinyurl.com/4e6hym79
Scribd :
https://tinyurl.com/pa85huh6
DOI :
https://doi.org/10.5281/zenodo.10753590
Abstract :
Chatbot, the AI Powered Pharmaceutical
Classification Systems using LLMs, can assist users in
efficiently navigating and understanding complex drug
classification. This research introduces the chatbot
framework empowered by LLM, to provide users with a
conversational user interface (UI) for Pharmaceutical
Classification inquiries.
The training of chatbot happened on a diverse
dataset, enabling it to grasp the intricate relationship
between drugs, dosage form, product type, pack size etc.
Through continuous interactions, the chatbot leverages
its contextual understanding to deliver real-time and
accurate information to users, ranging from healthcare
professionals seeking specific drug classifications to
consumers inquiring about medication details.
The research involves fine-tuning the pre-trained
language model such as Palm2, Llama2 and Meditron,
T5, Mistral 7B, TAPEX, BERT on a curated dataset of
drug related texts to enhance its understanding of
pharmaceutical concepts, molecular structure etc. The
fine-tuned model is then employed to classify drugs
based on multiple criteria including mechanism of
action, therapeutic class etc. The model’s ability to
comprehend complex relationships and contextual
information enables it to make accurate predictions and
handle ambiguous cases.
The Practical implication of this research extends to
pharmaceutical education, healthcare decision support,
and public health awareness. By offering a user -
friendly and conversational interface, the chatbot
provides an accessible and efficient means for
individuals across diverse backgrounds to obtain reliable
drug classification information. The study underscores
the transformative potential of LLMs in developing
intelligent chatbot tailored for pharmaceutical
knowledge dissemination, thereby contributing to the
evolving landscape of healthcare informatics.
Keywords :
Artificial Intelligence, Large Language Models, Chatbot Framework, Drug Classification, ,Conversational Interface, Real-Time Information.
Chatbot, the AI Powered Pharmaceutical
Classification Systems using LLMs, can assist users in
efficiently navigating and understanding complex drug
classification. This research introduces the chatbot
framework empowered by LLM, to provide users with a
conversational user interface (UI) for Pharmaceutical
Classification inquiries.
The training of chatbot happened on a diverse
dataset, enabling it to grasp the intricate relationship
between drugs, dosage form, product type, pack size etc.
Through continuous interactions, the chatbot leverages
its contextual understanding to deliver real-time and
accurate information to users, ranging from healthcare
professionals seeking specific drug classifications to
consumers inquiring about medication details.
The research involves fine-tuning the pre-trained
language model such as Palm2, Llama2 and Meditron,
T5, Mistral 7B, TAPEX, BERT on a curated dataset of
drug related texts to enhance its understanding of
pharmaceutical concepts, molecular structure etc. The
fine-tuned model is then employed to classify drugs
based on multiple criteria including mechanism of
action, therapeutic class etc. The model’s ability to
comprehend complex relationships and contextual
information enables it to make accurate predictions and
handle ambiguous cases.
The Practical implication of this research extends to
pharmaceutical education, healthcare decision support,
and public health awareness. By offering a user -
friendly and conversational interface, the chatbot
provides an accessible and efficient means for
individuals across diverse backgrounds to obtain reliable
drug classification information. The study underscores
the transformative potential of LLMs in developing
intelligent chatbot tailored for pharmaceutical
knowledge dissemination, thereby contributing to the
evolving landscape of healthcare informatics.
Keywords :
Artificial Intelligence, Large Language Models, Chatbot Framework, Drug Classification, ,Conversational Interface, Real-Time Information.
