The summary output produced by the online or off-line production of NA48 will contain indexing information for accessing the raw-data. We will provide a mechanism, transparent to the user, for retrieving raw-data for selected records by using this index. We shall also enable the user to build second-level indices by selecting records from the summary files. The user filters out the information of interest at the highest possible level and extends his knowledge about these objects by stepping down the hierarchy (user index, summary files, raw-data). Thus, full information has to be retrieved only for a small subset of the data.