Background: Data mining is characterized as a process of transforming data information into a humancomprehensible code format such as rules, formula, algorithm, and so on. Bioinformatics is developed to solve a \nbiological problem by using data mining technique. Identifying biomedical domain entities is a difficult task and the\nenhanced model is used to classify the entities from biomedical literature full text articles in PubMed database.\nMethods: The enhanced model involves 3 stages namely pre-processing, identification of the entities using dictionarybased approach and verification and validation with benchmarking databases. In Dictionary based approach, Disgenet \nand Pubtator are considered as the dictionary which is a benchmarking database. This approach defines entities using \nthe en_ner_bionlp13cg_md model from spacy package. \nResults and Conclusion: For experimental purposes, 99 full text articles related to Alzheimer\'s disease are considered \nwhich are downloaded from NCBI. Finally, demonstrated that our enhanced model for dictionary based approach \noutperforms in aspects of accuracy, precision and retrieval value. The enhanced model achieved 82% of accuracy \noverall. Compared to state-of -art method, the model obtained the better accuracy. These results suggested that the \nenhanced model is obtained high performance for extracting biomedical entities from PubMed articles. The \nimprovement is mostly due to the dictionary because Disgenet and Pubtator are considered the dictionary in the \nenhanced work. Since, the method is more fitting to classify biomedical entities.

Login to Download PDF

Please login with your Paper ID and password to access the full PDF.

🔑 Login to Download