Using Categories to Provide Context for Full-Text Retrieval Results

Using Categories to Provide Context for Full-Text Retrieval Results
复制标题

使用类别为全文检索结果提供上下文

DOI:
10.5555/2856823.2856835
复制
发表时间:
1994
期刊:
RIAO Conference
影响因子:
--
通讯作者:
Marti A. Hearst
Marti A. Hearst
中科院分区:
--
文献类型:
--
作者:
Marti A. Hearst

文献摘要

被引文献

相似文献

我们解决了一些问题,有关显示的结果,从全文检索。我们声称,显示查询结果的文档间的相似性是不合适的长文本,而不是分配类别,对应于文档的主要主题。我们认为,长文本的主要议题应该由多个类别表示,因为在许多情况下,一个类别不能充分分类的文本。我们描述了一个新的自动分类算法,不需要预先标记的文本和原型浏览界面,提出了一个简单的机制,显示多维信息。
We address some issues relating to display of results of passage retrieval from full-text collections. We claim that displaying query results in terms of inter-document similarity is inappropriate with long texts, and suggest instead assigning categories that correspond to documents' main topics. We argue that main topics of long texts should be represented by multiple categories, since in many cases one category cannot adequately classify a text. We describe a new automatic categorization algorithm that does not require pre-labeled texts and a prototype browsing interface that presents a simple mechanism for displaying multi-dimensional information.