Query Processing Based on Associated Semantic Context Inference

Abstract

Context-based query processing methods are used to capture user intents behind query inputs. General context models are not flexible or explicable enough for inference, because they are either static or implicit. This paper improves current context model and proposes a novel query processing approach based on associated semantic context inference. In our approach, the formal defined context is explicit, which is convenient to explore potential information during query processing. Furthermore, the context is dynamically constructed and further modified according to specific query tasks, which ensures the precision of context inference. For given query inputs, the approach builds concrete context models and refines queries based on semantic context inference. Finally, queries are translated into SPARQL for query engine. The experiment shows that the proposed approach can further improve query intents understanding to guarantee precision and recall in retrieval.