Buckets: Aggregative, Intelligent Agents for Publishing

Buckets are an aggregative, intelligent construct for publishing in digital libraries. The goal of research projects is to produce information. This information is often instantiated in several forms, differentiated by semantic types (report, software, video, datasets, etc.). A given semantic type can be further differentiated by syntactic representations as well (PostScript version, PDF version, Word version, etc.). Although the information was created together and subtle relationships can exist between them, different semantic instantiations are generally segregated along currently obsolete media boundaries. Reports placed in report archives, software might go into a software archive, but most of the data and supporting materials are likely to be kept in informal personal archives or discarded altogether. Buckets provide an archive-independent container construct in which all related semantic and syntactic data types and objects can be logically grouped together, archived, and manipulated as a single object. Furthermore, buckets are active archival objects and can communicate with each other, people, or arbitrary network services. KEYWO R DS: Digital library architectures, agents, archiving, multi-format, bucket, container, package. INTRODUCTION Digital libraries (DLs) are an important research topic in many scientific communities and have already become an integral part of the research process. However, access to these DLs is not as easy as users would like. Digital libraries are partitioned both by the discipline they serve (computer science, aeronautics, physics, etc.) and by the format of their holdings (technical reports, video, software, etc.). There are two significant problems with current DLs. First, interdisciplinary research is difficult because the collective knowledge of each discipline is stored in incompatible DLs that are known only to the specialists in the subject. The second significant problem is that although scientific and technical information (STI) consists of manuscripts, software, datasets, etc., the manuscript receives the majority of attention, and the other components are often discarded (Figure 1) [20]. Although nonmanuscript digital libraries such as the software archive Nettib [2] have been in use for some time, they still place the burden of ST! reintegration on the customer. A NASA study found that customers desire to have the entire set of manuscripts, software, data, etc. available in one place [19]. With the increasing availability of all-digital storage _xt transmission, maintaining the tight integration of the original STI collection is now possible. Old Dominion University and NASA Langley Research Center are developing NCSTRL+ to address the multidiscipline and multi-genre problems. NCSTRL+ is based on the Networked Computer Science Technical Report Library (NCSTRL) [5], which is a highly successful digital library offering access to over 100 university departments and laboratories since 1994, and is implemented using the Dienst protocol [9]. During the development stage, NCSTRL+ includes selected holdings from the NASA Technical Report Server (NTRS) [14] and NCSTRL, providing clusters of collections along the dimension of disciplines such as aeronautics, space science, mathematics, computer science, and physics, as well as clusters along the dimension of publishing organization and genre, such as project reports, journal articles, theses, etc. The DL aspects of NCSTRL+ are discussed in [ 15, 16]. Although developed for NCSTRL+ and with our modified version of the Dienst protocol in mind, buckets are protocol and archive independent, needing only standard World Wide Web (WWW) capability to function. This paper gives an overview of bucket functionality, examines similar work, and discusses current implementation and future plans. OVERVIEW Buckets are object-oriented container constructs in which logically grouped items can be collected, stored, and transported as a single unit. For example, a typical research project at NASA Langley Research Center produces information tuples: raw data, reduced data, manuscripts, notes, software, images, video, etc. Normally, only the report part of this information tuple is officially published and tracked. The report might reference on-line resources, or even include a CD-ROM, but these items are likely to be lost or degrade over time. Some portions such as software, can go into separate archives (i.e., COSMIC or the Langley Software Server) but this leaves the researcher to re-integrate the information tuple by selecting pieces from multiple archives. Most often, the software and other items, such as datasets are simply discarded. After l0 years, the manuscript is almost surely the only surviving artifact of the information tuple. Project manuscript _ library