The most important source of ingredients in the discovery of new drugs are Natural products. Moreover Nagoya protocol is most commonly used in selection of herbs based on similar efficiency, Later scientists have voiced their concern on protocol also proved it as less effective therefore, this project uses data mining classification approaches, novel targeted Selection which makes use of MED - LINE(Medical Literature Analysis and Retrieval system online) database that consists of biomedical information to identify herbs of same efficacy .Neural network technique among all classification techniques is inspired by biological nervous system. AS neural network is successful on wide array of noisy object selection of herbs is done effectively. SOM (self-organizing map) is most popular Neural Network provides a topology preserving mapping from the high dimensional space to map units. The main objective of this project is to survey on various data mining methods and their techniques and to conclude the suitable algorithm.