DOI倡议：元数据影响的共识探讨

需积分: 10 16 浏览量更新于2024-08-02 收藏 193KB PDF 举报

"《数字对象标识符（DOI）倡议：元数据影响分析》本文档旨在为国际DOI基金会生成一份关于元数据的共识论文，作为之前DOI现行立场共识论文的补充文献，参考了DOI1的相关内容。这份文档是对正在进行中的工作的总结，可能会随时进行修订。我们鼓励读者提出评论和批评，以便进一步改进。此版本是在先前版本的基础上编撰的，后续可能发布的修订版会考虑收到的意见，并可能包含政策调整或新的发展方向。请注意，此文档并不代表国际DOI基金会的永久性政策承诺，仅用于征求意见，完整复制时需保留版权声明。在DOI倡议中，元数据扮演着至关重要的角色。元数据是描述和定位数字对象的关键信息，包括但不限于资源的标题、作者、出版日期、版本、来源、类型、权利信息等。在DOI系统中，精确的元数据使得识别、引用和追踪网络上的学术、科技、商业等各种类型的内容成为可能。有效的元数据管理有助于确保跨平台的唯一性和可访问性，促进了知识共享和学术交流。 DOI的元数据标准规范了如何创建、管理和分配这些标识符，确保它们在全球范围内的一致性和互操作性。这包括使用特定的格式（如10.x.y.z/abcdefg形式），以及与URI结合，形成一个持久且可寻址的标识符。通过DOI，用户可以快速获取到所需资源的准确信息，从而简化了引用和检索过程，提高了学术研究的效率。然而，随着技术的发展和新的使用场景的出现，DOI倡议也在不断更新其元数据策略，例如考虑到Web 2.0环境下用户的生成内容、大数据分析的需求，以及与新的识别技术（如区块链）的融合。因此，元数据的设计和管理需要灵活适应变化，同时保持对隐私和数据安全的关注。总结来说，本论文探讨了DOI项目中元数据的重要作用，包括标准化、可追踪性和数据交换的便利性。它提供了关于如何处理和优化元数据以支持DOI体系的实用建议，同时也强调了持续改进和开放讨论的重要性，以应对未来挑战和需求。"

The DOI initiative: metadata implications Version 3: February 1999

The IDF is an affiliate of the INDECS activity and is using this as a platform to progress its

metadata development. INDECS has significant support from a wide range of the ADOI

community@; it is working on a tight schedule; it is practically oriented; it is run by experienced

and trusted individuals with a proven experience in building practical implementations of

metadata. It is important to recognise that INDECS is an example of the well-formed approach

used implicitly or explicitly in a number of other metadata activities, notably IFLA=s Functional

Requirements for Bibliographic Records [IFLA] and the Common Information System for

musical and other works [CIS]. Further, it has reached a consensus with the Dublin Core

community on the same approach [Bearman].

The “well-formed” approach is to define entities on the basis of their inherent structure, rather

than their application roles (e.g. place, person, creation, right, event, agreement); descriptive

elements of entities (e.g. for creations, elements might include the identifier or label itself;

controlled vocabulary descriptor labels, event, extent, relation) and types within these elements

(e.g. within the element Aextent@ would be the types Aspatial@ and Atemporal@). One of several

examples of existing metadata models following this Awell-formed@ approach is IFLA’s Functional

Requirements for Bibliographic Records [IFLA].

4.1.3 DOI essentials for metadata

From the current policies and future ambitions set out in section 1, and the acceptance of the

well-formed metadata principle, we can derive the following:

# Any DOI may turn up in any application;

# Applications must have a predictable answer to the question “what does this DOI

identify?”

# The only available point of quality control for this is at the point of allocation;

# DOI key metadata must therefore be:

# Declared on allocation

# Built from accessible vocabularies

# Accessible itself

# Adequate for identification

# Predictable to applications

# Extensible for local needs

# Persistent

# Authorised (the question of authority is discussed further below).

4.1.4 Metadata need not be centralised

The Internet creates a situation where metadata may be created once and remain in its original

location and yet be accessed by everyone (it is not essential that the metadata be stored in one

central location, merely that it be readily accessible and somehow authorised). The Internet Scout

Project [Roszkowski] for example is attempting to use distributed metadata for resource

The DOI initiative: metadata implications Version 3: February 1999

discovery. Equally, we want to minimise the size of the DOI core set and the work needed to

compile it (reducing unnecessary duplication). So we want the input which creates the kernel to

be done once, in a well-defined way (“do it once, do it right”), which does not reflect a specific

function but allows multiple functions. This can be done by designing the kernel metadata to

reflect the inherent structure of entities and metadata, rather than their function in any one context

or application [Rust]. Since the only available point of quality control for this is at the point of

allocation, there is a strong argument that the kernel DOI metadata must be either centrally

registered or registered via a pointer to a validated and authorised source of metadata.

4.2 Conformance with existing or developing standards

4.2.1 Standards for set of elements: the use of INDECS

There have been many efforts to establish metadata structures for particular communities;

examples include MARC records for libraries [USMARC], Federal Geographic Data Committee

for geospatial data [FGDC], Consortium for the Computer Interchange of Museum Information

[CIMI] etc. These provide a tool for those communities, but in the absence of a common data

model, interoperability between these Asilo@ applications is very limited and requires the

construction of specific bi-lateral Acrosswalks@ (mapping transformations between the unrelated

data models of each silo), with increasing complexity as the number of silos grows [StPierre]. In

a digital environment, this approach is no longer satisfactory as an infinite number of permutations

can easily be assembled (and need to be): an author of a scientific article, for example, may well

wish to insert a video clip, an animation, and a 3-dimensional structure model into his paper and

this must be readily accessible without the need to stop and define a specific Acrosswalk@ to link

the video clip metadata to the text metadata; the links must be automated and immediately usable.

One attempt to overcome this problem began by assuming that a common requirement for

interoperable metadata would be resource discovery, and has gained significant momentum and

recognition: the Dublin Core [DublinCore] initiative has now defined a core set of metadata for

the single functional purpose of resource description: a 15-element metadata element set intended

to facilitate discovery of electronic resources. There are significant problems with the DC

Asimple@ core set which limit interoperability (e.g. [Baker]) but recent consensus discussions

between of the Dublin Core, DOI, INDECS, and bibliographic standards communities (Oct 30/31

1998) have provided a way forward. It now appear that a set of elements capable of interoperable

use will be defined by this large community and receive widespread support [Bearman].

This offers for the first time the possibility of an activity which combines the widespread

community participation of the Dublin Core with the well-formed structured approach of

INDECS and others (e.g. IFLA, CIS). The DOI community is poised to be able to take

advantage of this. If this happens, the support of the DOI added to the existing participants

would give an overwhelming momentum to such a standard and its likelihood of adoption and

success. A stated aim of INDECS is to take its conclusions and proposed structures through the

relevant standardisation processes with the aim of these becoming formal recommendations.

剩余48页未读，继续阅读

paperopen

粉丝: 0
资源: 1

DOI倡议：元数据影响的共识探讨

海康威视mib

SNMP V2-mib的编写示例

CSharp Identifier Scraper:从 .cs 文件输出 C# 标识符名称的小工具-开源

mdadm: unrecognised metadata identifier: 0.95\

unbound identifier in: bitmap

qml中 readonly property identifier PI: 3.14159是什么意思

token identifier system : alter system kill session '1,2'

error C2065: 'Maxsize' : undeclared identifier C:\practice\1.cpp(81) : error C2061: syntax error : identifier 'Elemtype'

最新资源