Figure 2 provides the names and brief working definitions of the 10 Principal Elements of XOBIS, any one of which may serve as the nucleus of a particular Record.Each is discussed separately following the Attributes section.
Concept |
Topical and/or categorical constructs (tangible or intangible) not otherwise instantiated |
|
String |
Individual or deliberately clustered words or phrases, including numbers, letters, etc. |
|
Language |
Specific spoken, written, or signed communication systems |
|
Organization |
Organized groups, including jurisdictional subdivisions |
|
Event |
Named macro-events, naturally-occurring or conducted by individuals or organizations |
|
Time |
Individual chronological values or ranges of values (periods) |
|
Place |
Structures, geographic locations, and jurisdictions, including extraterrestrial ones |
|
Being |
Specific identities of tangible or intangible beings (living or dead) and/or personifications |
|
Object |
Manufactured, crafted, or naturally-occurring things, excluding Place, Being, and Work carriers |
|
Work |
Artistic or intellectual creations, excluding those considered Place or Object |
Figure 2. XOBIS Principal Elements
These 10 Principal Elements were isolated on the basis of their shared, homogeneous characteristics. This exercise was a lesson in linguistics and puzzle theory. The complex, frustrating, and sometimes-tortuous process of "slicing the pie" into fundamental categories appears to have achieved a coherent whole.The resulting structural consistency provides a foundation for potential optimization of indexing and for resolution of cataloging problems mentioned above.Since this document describes the alpha version of XOBIS, further review and practical application may necessitate redefinition or realignment of the various constituent elements. There are also some areas with potential for further simplification.As with any scheme, the need for associated rules or guidelines and the application of judgment will remain.However, XOBIS may potentially simplify these tasks.
The following prerequisites guided determination of elements in general.Retrieval ramifications informed the choice of necessary elements. Granularity of retrieval parallels granularity of definition; indeed a fundamental purpose of XML is to describe content, precisely for this reason. Similarly, while XML separates content from presentation, distinctions in display rely on mark-up existing at the desired level of detail.
A third goal in the delineation of Principal Elements was a desire to support enhancement of the "browse" capability, one which benefits greatly from libraries' authority control efforts, but which is limited or lacking entirely in Web-oriented search engines. Careful dissection and consideration of each of the Principal Elements resulted in identification of shared characteristics, which could underpin envisioned structured indexes, discussed below.
Lastly, selection was guided by consideration of the definition and categorization of relationships between elements. For example, real and fictional people share the same genealogical relationships. Subjects were problematical until we cast these as relationships between Principal Elements, e.g. a Concept being the topic of a Work, or a Being serving as the subject of a Work. The Relationships section elaborates.
Ten Conceptual Records are envisioned to record the definition and scope of each Principal Element and its relationships within XOBIS.At this time, the diagram in Figure 3 best explains the rationale behind decisions relating to the choice of Principal Elements.
Figure 3. XOBIS Core Structural Relationships
The Principal Elements represent varying degrees of specificity/generality of concepts, broadly interpreted, with the aim of being comprehensive. Tangibility and grammatical factors, especially proper nouns, were major determinants.Each Principal Element may be considered as representing a selected subset of instantiations (real or fictional) of a collective Concept element, discussed further below.They roughly represent instances of "substantive" or "notional" concepts.
Place, Being, Object, and Work (considering its physical carrier or container) are "substantive" in that they often represent the tangible, and are potentially collectable or ownable entities. This is reflected by a 'role' attribute with the value of instance, while the value authority permits inclusion of unheld and/or parallel intangible cases to allow referencing them consistently. The value authority/instance allows the same record to serve either purpose. The arrangement is such that the dowager Empress Cixi, the Eiffel Tower, the Hope Diamond, and the Gutenberg Bible can populate virtual collections of beings, places, objects, and works, just as readily as mundane specimens are more likely to populate physical collections. The same record may serve either or both purposes. An optional Version element accommodates coverage of versions (e.g. print versus digital) on a single record and may have discrete relationships. Holdings, a projected separate schema, would link directly to these Principal Elements or to one of their versions. Item, also a separate schema, would link to Holdings.
String, Language, Organization, Event, and Time are "notional" in that they represent ideations or intangibles, separated from the mostly generic Concept. Exemplars of these are all authorities. String permits extension of authority control to keywords in order to support eventual improvement of this most popular type of searching.Juxtaposing this approach with controlled topical vocabularies should prove interesting. Concept includes remaining abstract concepts, tangibles in the collective sense, e.g. the general idea of the heart as an anatomical pump occurring ubiquitously, as opposed to a single preserved cardiac specimen (Object), as well as specific intangible concepts.
XOBIS' integrative approach adds greater precision and flexibility to the current bibliographic apparatus, better accommodating the limitless variety of information found in digital environments, and making the schema inclusive of collections found in museums and related institutions.Arbitrary compartmentalization of information is increasingly a disservice to users, whose needs do not necessarily follow disciplinary or institutional boundaries. Harmonization of various practices might yield to various agencies' differential application of depth of XML hierarchies based on their specialization, perceived need, or available resources¯yet within the same structure.We hope that the recent establishment of the Institute of Museum and Library Services in the United States is fortuitous in promoting increased cooperation (42). The discussions of each Principal Element below include examples to illustrate their variety and to help clarify the rationale for their boundaries.These are preceded by coverage of Generic Elements and Attributes sections.