SGER: Semantic Technologies for Automatic Indexing and Tagging

从语义的角度出发，通过自动内容抽象来探索自动索引领域。稳健的语言处理技术用于生成多用途的数据结构，称为词汇网。它们依次用于自动语义索引程序，该程序设计用于解析、索引和超链接在线文本。研究主要涉及三个方面：词汇网实现的数据结构设计；设计词汇网络获取工具，利用语言学和统计启发式来开发强大的有限状态语法，用于重要的短语级和篇章级成分的分块；并设计文档注释工具，通过创建适当的超链接对带有词汇网的文本进行注释。词汇网的创建可以以高度索引和交叉索引的格式加速信息的可用性，促进导航和快速信息访问。

The area of automatic indexing is to be explored via automatic content abstraction from a semantic perspective. Robust language processing techniques are used to generate multipurpose data structures called lexical webs. These are, in turn, used in an automated semantic indexing program designed to parse, index, and hyperlink on-line texts. The research involves three major areas: data structure design for implementation of lexical webs; design of lexical web acquisition tools that utilize linguistic and statistical heuristics for developing robust finite-state grammars for chunking of significant phrase-level and discourse-level constituents; and design of document annotation tools that annotate a text with lexical webs by creating appropriate hyperlinks. The creation of lexical webs can accelerate the availability of information in a highly indexed and cross- indexed format, facilitating navigation and rapid information access.