Knowledge　graph　embedding　plays　a　vital　role　in　various　internet　of　things　applications,　such　as　knowledge　representation,　reasoning,　and　data　mining.　However,　existing　embedding　models　fall　short　in　effectively　utilizing　data　from　diverse　knowledge　domains　within　the　context　of　multiple　cross-domain　knowledge　graphs　while　ensuring　the　privacy　of　exchanged　data.　How　to　achieve　cross-domain　knowledge　embedding　in　the　integration　and　sharing　of　knowledge　information　is　an　urgent　problem　to　be　solved.　To　challenge　this　problem,　a　Cross-Domain　Privacy-Preserving　Knowledge　Graph　Embedding　(CDPPKGE)　algorithm　is　described　to　achieve　trusted　sharing　of　knowledge　information.　In　the　algorithm,　the　federated　learning　idea　is　introduced　to　realize　cross-domain　privacy　information　exchange　of　Knowledge　graph.　Specifically,　the　information　exchange　between　Knowledge　graphs　involves　two　stages:　Firstly,　the　RSA　encryption　mechanism　is　adopted　to　achieve　trusted　transmission　of　private　data　during　the　alignment　process　of　graph　entities.　Secondly,　differential　privacy　method　is　adopted　to　protect　the　trusted　transmission　of　private　data　in　the　process　of　cross-domain　embedding　alignment.　And　then　experimental　results　demonstrate　that　our　algorithm　achieves　better　graph　embedding　quality　while　preserving　the　privacy　of　sensitive　data.　©　2024　Copyright　held　by　the　owner/author(s).　Publication　rights　licensed　to　ACM.