An　intelligent　evaluation　method　was　proposed　for　the　credibility　of　answers　in　online　medical　communities　(OMCs).　By　applying　the　method　to　evaluate　and　classify　the　credibility　of　answers,　this　paper　aimed　to　inspire　users　to　adopt　reliable　health　information,　enhance　the　credibility　of　content,　and　support　OMCs　healthy　development.　The　study　constructed　a　content　knowledge　graph　for　answers　in　OMCs　and　a　domain　knowledge　graph　for　diabetes.　The　concepts　of　entity　regularity,　relationship　consistency　coefficient,　and　relationship　accuracy　were　introduced　to　calculate　credibility　scores　for　the　triples　in　the　community　answers,　which　would　be　aggregated　to　evaluate　the　content　credibility.　Validation　results　from　the　xywy.com　website　show　that　our　method　effectively　evaluated　and　classified　the　credibility　of　Q&A　content,　achieving　intelligent　identification　and　filtering　of　suspicious　answers.　The　precision　accuracy　of　credible　answers　is　92.5%,　significantly　improving　efficiency　and　interpretability　compared　to　manual　scoring　methods.　This　study　optimized　the　current　content　review　model　in　OMCs,　enhanced　content　management　efficiency　and　accuracy,　and　provided　feasible　tools　and　methods　for　monitoring　the　information　quality　in　OMCs　and　delivering　reliable　medical　knowledge　services.　©　The　Author(s),　under　exclusive　license　to　Springer　Nature　Singapore　Pte　Ltd.　2025.