Abstract
Structured knowledge bases are an increasingly important way for storing and retrieving information. Within such knowledge bases, an important search task is finding similar entities based on one or more example entities. We present QBEES, a novel framework for defining entity similarity based on structural features, so-called aspects and maximal aspects of the entities, that naturally model potential interest profiles of a user submitting an ambiguous query. Our approach based on maximal aspects provides natural diversity awareness and includes query-dependent and query-independent entity ranking components. We present evaluation results with a number of existing entity list completion benchmarks, comparing to several state-of-the-art baselines.
Highlights
Nowadays, more and more data becomes available in semantic form, be it, e.g., within the Linked Open Data (LOD) cloud (Heath and Bizer 2011), product databases or huge common knowledge bases like DBpedia (Auer et al 2007) or YAGO (Hoffart et al 2013)
While our model is general enough to be applied to the family of “Query by Entity ExampleS” (QBEES) use cases, we focus on list completion, since this is a well-defined information retrieval (IR) task with standard evaluation test-data available
While we focus on entity similarity search, our aspect model might be useful for modeling user interests in other application domains, including the entity summarization problem
Summary
More and more data becomes available in semantic form, be it, e.g., within the Linked Open Data (LOD) cloud (Heath and Bizer 2011), product databases (as used e.g. in shops like Amazon etc.) or huge common knowledge bases (ontologies) like DBpedia (Auer et al 2007) or YAGO (Hoffart et al 2013). Other applications include recommendation systems and general purpose entity search engines that provide similar entities given one or several examples With such an engine, a user might look (interactively) for movie directors that are actors, providing Quentin Tarantino and Clint Eastwood the user might expect to find Sylvester Stallone and Peter Jackson. Basic aspects represent single facets while our model explores the combinatorial facet space generating compound aspects to find the most similar entities under different points of view by identifying all entities that share a maximal set of semantic properties with the query. Contributions The main contributions of this paper are 1) the discussion in detail of our aspect-based entity model with its natural potential for modeling user interests and diversity awareness, 2) its extension by ad-hoc aspect relaxation, and 3) an evaluation of the approach.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
