论文标题
关于叙事信息访问数字库的合理性的维度
On Dimensions of Plausibility for Narrative Information Access to Digital Libraries
论文作者
论文摘要
设计基于关键字的访问路径是数字库中的常见实践。用户易于使用和接受，并为内容提供商带来适度的成本。但是，如果用户搜索感兴趣的故事比搜索一些关键字更复杂的故事，通常必须将搜索分解成零件。在搜索每一部分之后，必须手动重新组装信息。在先前的工作中，我们建议叙事信息访问，即，用户可以精确地将其信息需求作为图形模式称为叙述。然后，系统采用叙述，并为其每个部分寻找证据。如果整个查询（即每个部分）都可以与数据绑定，则叙述被认为是合理的，因此可以回答查询。但是那样容易吗？在这项工作中，我们进行案例研究以分析使给定叙事合理的过程。因此，我们总结了面对的概念问题和挑战。此外，我们在实现数字库中的叙事信息访问时必须考虑一组维度。
Designing keyword-based access paths is a common practice in digital libraries. They are easy to use and accepted by users and come with moderate costs for content providers. However, users usually have to break down the search into pieces if they search for stories of interest that are more complex than searching for a few keywords. After searching for every piece one by one, information must then be reassembled manually. In previous work we recommended narrative information access, i.e., users can precisely state their information needs as graph patterns called narratives. Then a system takes a narrative and searches for evidence for each of its parts. If the whole query, i.e., every part, can be bound against data, the narrative is considered plausible and, thus, the query is answered. But is it as easy as that? In this work we perform case studies to analyze the process of making a given narrative plausible. Therefore, we summarize conceptual problems and challenges to face. Moreover, we contribute a set of dimensions that must be considered when realizing narrative information access in digital libraries.
