Text mining approach for WSI in QA systems through relation construction

The paper attempts to identify the exact sense of the word through a process of establishing relations between similar words in the sample ontology. The philosophy relates to analysing the variance in sense among similar words with the help of a data mining tool to arrive at the exact sense of the target word as part of a question answering system. The focus spreads to derive new information from data and create patterns across datasets. The semantic relation among the words is calculated by assigning weights to investigate the variance and land at the exact sense. The distribution factor accrued from the manipulation of assigned weights reveals the significance of training to arrive at a better grading. The performance evaluated in terms of precision and accuracy exhibits the suitability of the proposed approach for use in the real world applications.