Abstract

Text mining is critical for information retrieval, especially in the health-care field.\nThe results of text mining are based on noisy data and false positives from natural language\ntext. This is owing to inconsistencies in semantics, syntax, the scarcity of class-specific core\nterms, and the enormous complexity of the data. As a result, traditional term/based techniques\nare ineffective for clustering medical articles containing complicated semantics. As a result, a\nsemantic-relationship based indexing technique for categorising research publications in the\nhealthcare sector is proposed with MeSH ontology as a domain reference to extract the relevant\nrelationships between the document\'s terms.