Electronic publishing of collections of XML documents should be accompanied with an information retrieval system to allow the users to access parts of documents which satisfy their specific needs. In this sense, different XML query languages appeared (XQL, for example) to express in queries what parts have to be retrieved from the documents. To evaluate a query, several steps are performed with respective intermediate results until the desired final result. In this paper, we present aninteractive XML query language (called IXDIRQL) which is an extension to XQL to allow for the selection on the spot of the interesting subset of elements from each intermediate result of a query. To make this possible, intermediate results must be available during the construction of the query. This helps the user in constructing his query to achieve the desired result. Moreover, in IXDIRQL textual similarity search of traditional Information Retrieval is also possible. The result of this operation is a ranked list of elements organized in a way to ease the task of the user to detect the relevant information in the results.