arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1003.4418cs.DBcs.IR
收藏

实体搜索引擎查询生成器的评价

Evaluation of Query Generators for Entity Search Engines

Stefan Endrullis, Andreas Thor, Erhard Rahm

摘要:

动态Web应用程序(如混搭)需要高效地访问只能通过实体搜索引擎(如产品或出版物搜索引擎)访问的Web数据。然而,当前大多数mashup系统和应用程序只支持简单的关键字搜索,以便从搜索引擎检索数据。我们建议使用更强大的搜索策略建设所谓的查询生成器。对于给定的一组实体,查询生成器能够自动确定一组搜索查询以从实体搜索引擎检索这些实体。我们证明了查询生成器的有用性,按需Web数据集成和评估的有效性和效率的查询生成器的一个具有挑战性的现实世界的集成方案。

英文摘要:

Dynamic web applications such as mashups need efficient access to web data that is only accessible via entity search engines (e.g. product or publication search engines). However, most current mashup systems and applications only support simple keyword searches for retrieving data from search engines. We propose the use of more powerful search strategies building on so-called query generators. For a given set of entities query generators are able to automatically determine a set of search queries to retrieve these entities from an entity search engine. We demonstrate the usefulness of query generators for on-demand web data integration and evaluate the effectiveness and efficiency of query generators for a challenging real-world integration scenario.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑