Relational　databases　in　PLM　(product　lifecycle　management)　system　are　commonly　searched　using　structured　query　languages.　The　user　needs　to　know　the　data　schema　to　be　able　to　ask　suitable　queries.　In　contrast,　the　IR　(information　retrieval)　community　has　popularized　keyword-based　search　that　allows　users　to　use　keywords　to　access　unstructured　data.　In　this　paper,　we　integrate　IR　techniques　with　structure　data　search　and　present　an　approach　for　querying　PLM　databases　using　keywords.　This　approach　has　two　significant　novel　features:　an　ontology-based　query　expansion　and　dealing　with　semantic　uncertainty　of　keywords　based　on　extended　belief　network.　We　applied　this　approach　to　a　practical　PLM　database　to　verify　the　usefulness　of　the　proposed　method.　©　(2012)　Trans　Tech　Publications,　Switzerland.