Home About Login Current Archives Announcements Editorial Board
Submit Now For Authors Call for Submissions Statistics Contact
Home > Archives > Volume 20, No 8 (2022) > Article

DOI: 10.14704/nq.2022.20.8.NQ44816

A Named Object Recognition Model of Health Records Using Improved Cuckoo Search Optimization Algorithm and Kernel Extreme Learning Machine

R. Ramachandran, Dr. K. Arutchelvan


Named Object Recognition (NER) has become an important area in several natural language processing technologies such as information extraction and information retrieval. The benefits of NER have attracted more attention among researchers in diverse fields. But the class labels and extension of named entities considerably differ with respect to distinct application domains. Medical literature includes valuable details namely clinical signs, diagnosis, drug, and medication for particular diseases. As the knowledge gaining from health literature is a tedious task, this paper aims to develop a new NER model in medical literature using Improved Cuckoo search algorithm (ICSA) with a Kernel Extreme Learning Machine (KELM), called ICSA-KELM. The presented ICSA-KELM model involves three major stages namely preprocessing, classification, and parameter tuning. The presented model performs preprocessing to convert the raw medical data into a useful format. In addition, the KELM model is executed to perform a classification process. Finally, the two main parameters of KELM such as penalty parameter 'C' and kernel bandwidth 'γ' of the KELM are computed using the ICSA algorithm. An general set of recreations were carried out to determine the effectual classification performance of the ICSA-KELM model. The resultant experimental values ensured the betterment of the ICSA-KELM model over the existing methods.


Machine learning, Named entity recognition, Parameter tuning, Medical literature, KELM model

Full Text