A methodology for user charge estimation in numeric online databanks. Part II

Abstract

Recent studies on user charging information retrieval services emphasize charging on the basis of the information retneved from the databases or its retneval costs as being more justified and appropnate as bases for charging than the tradi tional connect-time. When these charging bases are applied in numeric online databanks, the charges vary over a very broad range, depending on factors which the users cannot know or manage. Therefore the users cannot estimate their charges in advance, prior to query execution. This has many undesirable practical consequences for information retrieval and informa tion seeking. In order to avoid such consequences and to achieve full advantage of such charging bases, user charge estimation tools which automatically estimate the charges on the basis of user query statements must be provided in the query systems to databanks. A novel and systematic methodol ogy for user charge estimation, to be applied by the tools, is reviewed. The methodology has been developed in the context of the relational data model. By applying the methodology, user charges for the information retrieved and its retneval costs for relational algebra queries can be estimated generally, accu rately and conveniently. In addition to the charge estimates, the methodology provides rich descriptions of the query results that essentially support query formulation. Therefore the user can concentrate on query formulation with a clear idea about the query results and his charges in mind. The architecture of the methodology is such that the methodology can be extended to further charging bases in information retrieval (e.g. data communication costs). In Part I of the article, the state of the art and trends in the development and use of databanks and in the charging principles of information retneval were reviewed. The methodology is now descnbed and special attention is paid to the design of user interfaces of the tools based on the methodology.