Educational Tools in Herbal Medicine: A Streamlit-Based AI Decision Tree Classifier for South Indian Medicinal Herb Identification ("PLANTIFY")

Educational Tools in Herbal Medicine: A Streamlit-Based AI Decision Tree Classifier for South Indian Medicinal Herb Identification ("PLANTIFY")

J.Jaydish , G.Chelladurai

Department of Botany, St. Joseph’s College (Autonomous), Tiruchirappalli, TamilNadu, India

Corresponding author Email:jaydishkennedy@gmail.com

Abstract:

The application of emerging artificial intelligence, particularly the decision tree classifier algorithm, enables the accurate identification and classification of plant species for herbal south Indian medicinal herbs that are vital to traditional systems like Ayurveda and Siddha. However, identifying these herbs is challenging due to their complex morphology and limited taxonomic resources. To address this, we developed PLANTIFY, a web-based app using streamlit and a decision tree classifier trained on eight key morphological traits. The model identifies 100 South Indian herbs with 92.5% accuracy using 5-fold cross-validation. The app provides species predictions with confidence scores, detailed taxonomy, ethnobotanical uses, and downloadable PDF reports. A usability study found 90% of users rated the app as user-friendly. PLANTIFY bridges traditional knowledge with AI, promoting herbal education and preserving ethnobotanical heritage. For research into and the identification of plant species for taxonomical purposes, this emerging technology is more convenient and innovative.

Keywords:

morphological trait; tensor flow; streamlit; Ayurveda; plant identification