Information retrieval is used to identify the relevant documents in a document collection, which is matching a user's query. It also refers to the automatic retrieval of documents from the large document corpus. The most important application of information retrieval system is search engine like Google, which identify those documents on the World Wide Web that are relevant to user queries. In most situations, users may download the files that are already downloaded and stored in their computer. Then, there is a chance of multiple copies of the files that are already stored in different drives and folders on the system, which in turn reduces the performance of the system and these files occupy a lot of memory space. Analyzing the contents of the file and finding their similarity is one of the major problems in text mining and information retrieval. The main objective of this research work is to analyze the file contents and deletes the duplicate files in the system. In order to perform this task, this research work proposes a new tool named Duplicate File Detector Tool i.e. DFDT. DFDT helps the user to search and delete duplicate files in the system at a minimum time. It also helps to delete the duplicate files not only with the same file category, but also with different file categories. Boyer Moore Horspool and Knuth Morris Pratt string searching algorithms are existing algorithms and these algorithms are used to compare the file contents for finding their similarity. This work also proposes a new string matching algorithm named as W2COM (Word to Word COMparison). From the experimental results it is observed that the newly proposed W2COM string matching algorithm performance is better than Boyer Moore Horspool and Knuth Morris Pratt algorithms.