An Intelligent Approach for Multi Disease Specific Plant Corpora Creation And Plant Name Extraction From Biomedical Literatures

Indian Medicinal plants play an important role in the treatment of various diseases disputedly most of the medicinal plants are available as an unstructured format in biomedical literatures. Since the growth of biomedical literature tripling every year despite there is no proper text mining tool to automatically extract the medicinal plants from the works of literature and to automatically create the disease-specific plant corpora. Our proposed algorithm utilizes four methods to automatically create the disease-specific plant corpus by extracting the plant names from the multi-disease biomedical literature automatically. The first method uses the multi-keyword indexing to create the multi-disease corpus by referring to the appropriate disease based biomedical articles and the second method uses the rule-based system based on python techniques to automatically extract the plant names and chemical compounds from the multi-disease based biomedical works of literature. The third method uses the fuzzy techniques to extract unique plants from multiple disease articles. The fourth method uses the Tf-Idf to rank the top- ranked plants extracted from the multiple diseases. Our proposed model is implemented for four diseases namely Cancer, Diabetes, Arthritis, and Fever. The proposed model is validated by using the human inputs and our model outperforms the state of art approach by scoring the Accuracy value of 95% average for Automatic disease corpus creation and Automatic plant extraction from biomedical works of literature abstracts and results.