Relevance Search over Schema-Rich Knowledge Graphs

• DOI: 10.1145/3289600.3290970
• Publication Date: 2019-03-11T12:33:01Z
• Authors: Yu Gu; Tianshuo Zhou; Gong Cheng; Ziyang Li; Jeff Z. Pan; Yuzhong Qu
• Keywords: Schema (genetic algorithms); Relevance; Heuristics; Knowledge graph

Relevance search over a knowledge graph (KG) has gained much research attention. Given query entity in KG, the problem is to find its most relevant entities. However, relevance function hidden and dynamic. Different users for different queries may consider from angles of semantics. The ambiguity more noticeable presence thousands types entities relations schema-rich which challenged effectiveness scalability existing methods. To meet challenge, our approach called RelSUE requests user provide small number answer as examples, then automatically learns likely these examples. Specifically, we assume intent can be characterized by set meta-paths at schema level. searches KG diversified significant that best characterize user-provided examples entity. It reduces large space using distance degree-based heuristics, performs reasoning deduplicate represent equivalent query-specific Finally, linear model learned predict meta-path based relevance. Extensive experiments demonstrate outperforms several state-of-the-art