Semantic information retrieval is a key application in the field of intelligent information retrieval. Under the condition of knowledge incompleteness, how to realize the semantic information retrieval is its foothold. By using rough ontology into semantic information retrieval, the rough ontology based semantic information retrieval model and its concrete realization scheme is put forward. Its concerned future work will be done from two aspects as below.
Firstly, since the annotation of rough ontology mainly adopts the semi-automatic method, instance identifying of the classes in rough ontology from documents still needs to be done, so the precise automatic annotation should be implemented to promote the efficiency of system ROSRS.
Secondly, because semantic extension supported by rough ontology may leads to an extreme expansion with the retrieved results, so the selection criteria of semantic similarity threshold and other related parameters need to be got through system self-study aided by domain expert, and to be autonomously adjusted in the trial.