With　the　rapid　development　of　urbanization,　a　series　of　burglary　cases　occurred　frequently.　It　is　very　important　to　push　suspicious　cases　according　to　the　top　rank　in　order　to　improve　the　ability　of　intelligent　concatenation　cases.　This　article　proposes　a　method　of　selecting　burglary　cases　based　on　multidimensional　features　and　PageRank,　which　could　analyze　the　fact　description　of　burglary　cases,　extract　the　multidimensional　features　of　cases　and　perform　feature　representation　to　obtain　the　feature　vector　of　each　case.　Then　it　constructed　the　network　through　the　similarity　between　the　multidimensional　features　and　used　PageRank　to　calculate　the　importance　of　each　case　for　ranking.　At　last,　the　algorithm　is　verified　based　on　the　real　data　of　burglary　cases.　The　experimental　results　showed　that　the　robustness　test　is　better　than　the　other　two　methods.　And　the　proposed　method　is　more　effective　and　can　be　better　applied　to　the　sorting　of　burglary　cases.