Ontology Learning in Text Mining for Handling Big Data in Healthcare Systems

The field of biomedical informatics has been growing fast over the past few decades. Due to rapid advancement in the form of digitized biomedical information, such as patient lab reports, patient medical records, prescription, and physician notes enormous amount of unstructured biomedical data are generating every day. However, extracting required information and sharing it in different application remain a challenging task. The categorization of unstructured text in biomedical science is one of the fundamental data analysis techniques that have been widely used for managing abundant textual data (Hsieh, S., et al., 2011. Enabling the development of base domain ontology through extraction of knowledge from engineering domain handbooks. Advanced Engineering Informatics, 2(25), pp.288–296). Ontology offers the potential for providing a logical interpretation of biomedical textual data that is based on a hierarchical conceptual representation of information. However, one of the major obstacles that prevents ontology from being deployed in large-scale biomedical information systems is ontology acquisition, which strongly depends on knowledge engineers and domain experts. Additionally, ontology building is a labor-intensive, handcrafted, and recursive process. Therefore, to address the above mentioned problem, researchers have devised semi-automatic techniques called ontology learning for building ontologies. This survey provides a comprehensive analysis of ontology learning techniques, such as linguistic, statistical, and semantic-based techniques, extensively used in ontology learning process. Moreover, the survey provides a detailed review of the ontology learning process in the field of biomedical systems.