A novel similarity measure technique for clustering using multiple viewpoint based method

Data mining is nothing but the process of automatically searching large stores of data to discover patterns and trends that go beyond simple analysis. So it is observed that while doing clustering there may be a chance of occurring dissimilar data object in a cluster. This paper introduces such technology that makes the patterns more accurate, and it helps to search more accurate analysis of data. This System greedily picks the next frequent item set in the next cluster. For this the multiple viewpoints are used to measure the similarity between two different data objects is introduced. We can define similarity between two objects explicitly or implicitly. Cosine similarity measures will resolve this problem. As multiple viewpoints will focuses on similarity measures at multiple levels. These criteria will be used to group the documents based on similarity. The similarity measured between current cluster documents and also other cluster group documents.