Systematic　Brain　Informatics　(BI)　depends　on　a　lot　of　prior　knowledge,　from　experimental　design　to　result　interpretation.　Scientific　literatures　are　a　kind　of　important　knowledge　source.　However,　it　is　difficult　for　researchers　to　find　really　useful　references　from　a　large　number　of　literatures.　This　paper　proposes　a　personalized　method　of　literature　recommendation　based　on　BI　provenances.　By　adopting　the　interest　retention　model,　user　models　can　be　built　based　on　the　Data-Brain　and　BI　provenances.　Furthermore,　semantic　similarity　is　added　into　traditional　literature　vector　modeling　for　obtaining　literature　models.　By　measuring　similarity　between　the　user　models　and　literature　models,　the　really　needed　literatures　can　be　obtained.　Results　of　experiments　show　that　the　proposed　method　can　effectively　realize　a　personalized　literature　recommendation　according　to　BI　researchers＇　interests.