As the data is growing everyday and comparatively less amount of useful information is made available on the internet, it becomes necessary to introduce a mechanism that can easily search out relevant information from that bulk of data. This is what has contributed toward the rise of the concept of multi-document summarization where the whole document is condensed to a smaller version retaining its original meaning. We are going to work on automated creation of summaries of one or more text documents using similarity measure algorithm – cosine similarity.