Discovering Context of Labeled Text Documents Using Context Similarity Coefficient

To find closeness between two data points, traditional distance based closeness measurement calculates distance between two data points. However, it fails to capture behaviour of data series. Behaviour of data series can be captured by association and disassociation between patterns of data points. This can reflect closeness between them. The same concept can be applied to find association between text documents. Using this philosophy, this paper proposes a novel approach of document association based on context similarity coe_cient (CSC). CSC based document association helps to capture contextual relationship between documents. Experiments conducted on standard datasets such as Reuters-21578 and RCV1 show that CSC successfully finds closeness between the documents.