Description

The development of ontologies involves continuous but relatively small modifications. Even after a number of changes, ontology and its previous version usually share most of their axioms. Unfortunately, when ontology evolves, current reasoners do not take advantage of the similarities between the ontology and its previous version. That is, when reasoning over the latest version of ontology, current reasoners do not reuse existing results already obtained for the previous one and repeat the whole reasoning process. For large and complex ontologies this may require a few minutes, or even a few hours. Cognitive on a Web scale becomes increasingly stimulating because of the large volume of data involved and the complexity of the task. Full re-reasoning over the entire dataset at every update is too time-consuming to be practical. Semantic information has been reducing by using hadoop framework with simple machine learning algorithm. Each level of mapping and reducing based on k-means clustering technique. Large set of information can be constructing or modify with the help of simple pattern based grouping. Dynamically grouping dependencies can be made based on attributes. Clustered values have get modification like addition. At the end user query has been retrieve with the help of grouped items. We have assessed our system on the BTC benchmark and the results show that our method outperforms related ones in nearly all aspects.