Semantic aggregation is a technique that has been used frequently in Natural Language Processing to generalize a group of entities by means of a category or group to which they belong to such as "mammals", "rodents","continents" etc. This technique amicably avoids repetitive words and phrases encapsulating them under a single term. Recently this methodology has received a lot of attention in automatic summarization. It has presented its challenge in finding appropriate aggregations in order to produce high quality summaries. In this project we would be exploiting the hyponomy \texttt{is-a} relationship in the form of Hearst pattern \cite{hearst1992automatic} to identify good quality aggregations to produce more general and simplified sentences that can ultimately generate superior summaries.