Domain-specific searching using concepts

在专利数据库或医疗集合等领域内进行搜索可能会很困难，无论是在编写有效的查询还是在评估结果方面。搜索者经常使用与文档集合中使用的术语不同的术语，并且许多不相关的文档可能包含在搜索结果中。在我们之前的研究中，我们探索了使用概念知识库来支持Web搜索者制作查询并随后探索最热门的搜索结果。本研究的目标是将基于概念的搜索框架从我们之前的研究扩展到特定领域集合内的搜索。最初的目标信息域在专利数据库中。将利用现有的专利分类构建一个新的概念知识库。先前开发的用于支持用户进行交互式查询细化和交互式搜索结果探索过程的技术，将为支持这些任务的接口开发的进一步研究奠定基础。一个主要的研究目标是进一步了解搜索者如何利用概念信息和表示这些信息的接口来满足他们的信息需求。本研究的预期成果包括：(1)一个描述专利知识的概念知识库；(2)使用概念知识库支持查询的交互式细化和搜索结果的交互式探索的原型系统；(3)完成原型系统的用户评估；(4)为保护本研究过程中产生的新知识产权而提出的专利申请；(5)通过会议报告和期刊出版物向公众传播研究成果。这项工作可以很容易地扩展到其他文档集合，如医疗数据库、公司文档集合等。

Searching within domains such as patent databases or medical collections can be difficult, in both crafting effective queries and in evaluating the results. Searchers often use terminology that is different from what is used within the document collection and many non-relevant documents may be included in the search results. In our previous research, we explored the use of a concept knowledge base to support Web searchers in crafting a query and subsequently exploring the top search results.The goal of this research is to extend the framework of concept-based searching from our previous research to searching within domain-specific collections. The initial target information domain is within patent databases. A new concept knowledge base will be constructed using existing classifications of patents. Previously developed techniques for supporting users in their interactive query refinement and interactive search results exploration processes will form the basis for further research in the development of interfaces to support these tasks. A primary research goal is to further understand how searchers can make use of the conceptual information and the interfaces for representing this information to fulfill their information needs.The anticipated outcomes of the completed research include: (1) a concept knowledge base describing patent knowledge; (2) a prototype system that supports the interactive refinement of queries and the interactive exploration of search results, using the concept knowledge base; (3) completed user evaluations of the prototype systems; (4) patent applications to protect any new intellectual property created in the course of this research; and (5) public dissemination of the research outcomes through conference presentations and journal publications. This work may be readily extended to other document collections such as medical databases, corporate document collections, ect.