Information Retrieval Based on a Query Document Using Maximal Frequent Sequences

Information Retrieval (IR) methods are commonly based on words, these methods allow the user to formulate a query through keywords. However, there are situations where the user has only one example document and based on this example it is needed to recover the most similar documents in a collection. This paper proposes an IR method that receives as input a query document and retrieves the k most similar documents to the query document using a representation based on Maximal Frequent Sequences (MFSs). Our method is tested and compared against the IR model based on bag of words, the experimental results show that the proposed method obtains good performance in contrast to the results obtained by the IR model based on bag of words.