WWW is composed of a great volume of documents that are stored by several data sources. Normally, a user is interested in those documents that include certain keywords. However, these documents might be incomplete, ancient or huge, and therefore, the user would have to discard irrelevant ones. An ideal Web search tool might select the best documents in base on user criteria defined over quality parameters such as completeness, recentness, frequency of updates and granularity. Traditional query languages are very restrictive in expressing preference-based queries. Therefore new query languages, such as SQLf, are needed. We present a tool that allows the selection of the best data sources and documents in terms of user preferences. Documents and data sources are described according to quality parameters. User preferences are expressed by means of SQLf queries. Our tool contains a wizard to retrieve the best documents and data sources. Thus, the user is oriented by a set of steps where a preference query that involves quality parameters is built easily.