Extracting most significant data about the user queries from the search engine by K-means++ algorithm

Today every process requires very important data as well as updated data about that process or work. In order to acquire those data, employees from various fields are searching through different search engines. Only very few times search engines are helping to get the users expected data but many times they are providing only the approximate data about the user expectation. To avoid this state, we used the algorithm named K-means++ over the search engine documents to list the most relevant information, because this algorithm uses the special mathematical method to find the out the successive cluster center of each cluster documents and only the first center is random selection from the data unlike K-means algorithm, which randomly selects all the cluster centers. Segmentation fusion is applied to provide the most resultant list, which is accepting as input of each cluster's documents gradually increasing manner. Performance of the K-means++ algorithm is compared and evaluated with the measures like Mean Average Precision and F-measure, which shows the novel algorithm has providing the better relevant result than other traditional algorithms. Finally, this algorithm also works well for large data sets. So, the researchers can utilize this algorithm to their innovative ideas comes to reality with the performance matrices of high relevancies such as precision and recall.