Traditional information retrieval (IR) systems evaluate user queries and retrieve/rank documents based on matching keywords in user queries with words in documents.These exact word-matching and ranking approaches ignore too many relevant documents that do not contain the exact keywords as specified in a user query. Instead of considering these traditional approaches, we propose to retrieve documents using a fuzzy set IR model and rank retrieved documents for any vague query using the "vagueness score" of the documents based on the word senses as defined in WordNet. Using the vagueness scores, we rank the most highest "relevant" documents of a vague query qas the ones that best cover the different possible senses of keywords in q. The proposed word-sense ranking method enhances the existing ranking approaches on ordering retrieved documents for vague queries and thus provides a more reliable and elegant tool for information retrieval.