Abstract: Various information retrieval methods are retrieve the prepared and correct information from large number of datasets, which are expensive to manage in terms time and money. Particularly when work on important text that may not contain any instance of the embattled planned information. The different approach that serves and also creates well arranged metadata using attributes which are reasonable by identify the data store that are likely to contain information of user interest or related to search keyword enter in query and this information is going to be consequently useful for querying and extracting from the database. We propose the method CADS (Collaborative Adaptive Data Shearing Platform) is based on the query modules and the metadata with clustering using which metadata added in datasets and search using the query forms in which resulted records are downloadable and readable. It represents the way to recognize structured information and extraction of information make easy. It is useful to improve the search efficiency by using the content search and query search with use of clustering searching is faster.

Keywords: Query, Annotation, CADS.