语言资源管理.特性结构.第2部分:特性系统声明

[中文适用范围]： ISO 24610-2 提供了一种格式，用于在自然语言应用中表示、存储或交换特征结构（feature structures），适用于语言数据的注释和生成。其最终目的是提供一种计算机格式，用于定义类型层次结构，并声明适用于特征规范集合及其对特征结构操作的约束，从而提供一种手段来检查每个特征结构是否符合参考规范的合规性。特征结构是许多语言学形式的重要组成部分，也是语言工程应用程序中消耗、生产和用于的底层机制。特征系统声明（FSD）是与使用特征结构（fs）元素的特定类型文本配合使用的辅助文件。FSD 具有以下四个目的：1. 提供一种编码方式，用于引入和定义类型及其子类型和继承关系，从而构建特征系统的基础；2. 提供一种机制，使编码器能够列出所有特征名称和特征值，并提供描述性文字说明其含义；3. 提供一种机制，用于声明类型约束，Typed 特征结构可依据这些约束相对于 Typed 特征逻辑中声明的特定理论进行验证。这些约束可能涉及特征值范围的约束、特定类型特征结构中允许的特征的约束，或防止某些特征值对共同出现的约束。约束的来源通常是待建模的经验领域；4. 提供一种机制，使编码器能够定义未指定特征结构的预期解释。这涉及为缺失的特征定义默认值（无论是字面默认值还是计算默认值）。本部分描述的方案可用于记录任何特征系统，但主要用于与 ISO 24610-1 中定义的 Typed 特征结构表示法配合使用。ISO 24610-1 的特征结构表示法规定了受 ISO 24610-2 中指定的类型约定和约束限制的数据结构。ISO 24610-1 的特征结构表示法也用于 ISO 24610-2 中定义的一些元素内。 [外文原描述]： ISO 24610-2:2011 provides a format to represent, store or exchange feature structures in natural language applications, for both annotation and production of linguistic data. It is ultimately designed to provide a computer format to define a type hierarchy and to declare the constraints that bear on a set of feature specifications and operations on feature structures, thus offering means to check the conformance of each feature structure with regards to a reference specification. Feature structures are an essential part of many linguistic formalisms as well as an underlying mechanism for representing the information consumed or produced by and for language engineering applications. A feature system declaration (FSD) is an auxiliary file used in conjunction with a certain type of text that makes use of fs (that is, feature structure) elements. The FSD serves four purposes. 1) It provides an encoding by which types and their subtyping and inheritance relationships can be introduced and defined, thus laying the basis for constructing a feature system. 2) It provides a mechanism by which the encoder can list all of the feature names and feature values and give a prose description as to what each represents. 3) It provides a mechanism by which type constraints can be declared, against which typed feature structures are validated relative to a given theory stated in typed feature logic. These constraints may involve constraints on the range of a feature's value, constraints on which features are permitted within certain types of feature structures, or constraints that prevent the co-occurrence of certain feature-value pairs. The source of these constraints is normally the empirical domain being modelled. 4) It provides a mechanism by which the encoder can define the intended interpretation of underspecified feature structures. This involves defining default values (whether literal or computed) for missing features. The scheme described in ISO 24610-2:2011 may be used to document any feature system, but is primarily intended for use with the typed feature structure representation defined in ISO 24610-1. The feature structure representations of ISO 24610-1 specify data structures that are subject to the typing conventions and constraints specified using ISO 24610-2:2011. The feature structure representations of ISO 24610-1 are also used within some of the elements defined in ISO 24610-2:2011.