In　order　to　meet　the　needs　of　ordinary　users＇　fuzzy　query,　and　to　take　into　account　of　the　complexity　and　precision　rate　of　implemented　system,　this　paper　summarizes　existed　user-query-oriented　theory　and　presents　a　method　for　fuzzy　query　based　on　PBFS　and　NQL　language.　This　paper　proposes　a　SQL　language　at　first　with　simple,　intuitive　features;　Secondly,　we　propose　a　PBFS　algorithm　which　can　automatically　generate　SPARQL　in　the　system　based　on　the　fuzzy　input,　so　as　to　obtain　accurate　results.　This　algorithm　has　low　complexity,　convenient　input,　high　accuracy　and　other　characteristics;　Finally,　this　schema　is　verified　through　the　experiments.　The　results　of　experiments　show　that　this　program　can　effectively　improve　the　user　search　efficiency　according　to　its　accurate,　complex　characteristics.　(C)　2016　The　Authors.　Published　by　Elsevier　B.V.