Structural Query Expansion via motifs from Wikipedia

Abstract

The search for relevant information can be very frustrating for users who, unintentionally, use inappropriate keywords to express their needs. Expansion techniques aim at transforming the users' queries by adding new terms, called expansion features, that better describe the real users' intent. We propose Structural Query Expansion (SQE), a method that relies on relevant structures found in knowledge bases (KBs) to extract the expansion features as opposed to the use of semantics. In the particular case of this paper, we use Wikipedia because it is probably the largest source of up-to-date information. SQE is capable of achieving more than 150% improvement over non-expanded queries and is able to identify the expansion features in less than 0.2 seconds in the worst-case scenario. SQE is designed as an orthogonal method that can be combined with other expansion techniques, such as pseudo-relevance feedback.