作者: Gong Cheng , Yuzhong Qu
关键词:
摘要: Along with the rapid growth of data Web, searching linked objects for information needs and reusing become emergent ordinary Web users developers, respectively. To meet challenge, we present Falcons Object Search, a keyword-based search engine objects. serve various keyword queries, each object system constructs comprehensive virtual document including not only associated literals but also textual descriptions links The resulting are ranked by considering both their relevance to query popularity. For object, query-relevant structured snippet is provided show matched query. Besides, Web-scale class-inclusion reasoning performed discover implicit typing information, could navigate class hierarchies incremental class-based results filtering. task-based experiment promising features system.