In the present world, due to tremendous development in technology, a huge amount of information is available everywhere. Therefore, it is difficult for the users to understand the main content of the entire document as it takes a lot of time. In this work we use extractive text summarization which uses a method to give the version of summary for one or more file or document. Here we give an approach that maps sentences to nodes of a hierarchical ontology. Ontology explains what exists in a particular domain. For the ontology creation, vocabularies are collected. It is used as background knowledge and helps to find the related meaning of the terms which occur in the source documents. Text mining is the technique from which high quality information is derived from text. Clustering is a significant task. The clustering method groups similar or related terms into a single group. In the first stage, data collection takes place. The preprocessing stage includes stemming and stop words removal.TF-IDF process occurs after which clustering takes place. In the ontology creation, first the determination of the main sub topics of the article of interest is done. We classify sentences to nodes which have a predefined hierarchical ontology. Each ontology node has bag-of-words from a web search. We represent sentences by sub trees that permit to apply measures of similarity and find relations between sentences. The ontology used in this work is not domainspecific; it does not require labelled data. this work can be extended to topics focused on summarization framework to news articles or blogs and to also to various machine learning approaches