To　achieve　environmental　false　complaint　report　detection,　a　false　complaint　and　report　detection　model　was　proposed　based　on　adversarial　transfer　learning　method.　First,　a　long-short　term　memory　(LSTM)　network　was　used　as　a　feature　extractor　to　extract　the　shared　features　of　Weibo　rumors　(source　domain)　and　complaint　report　text　(target　domain).　Second,　the　domain　adaptation　was　performed　by　using　the　adversarial　learning　method　to　align　the　source　domain　features　with　the　target　domain　features.　Finally,　the　classification　results　were　output　by　the　classifier,　and　the　network　parameters　were　updated　by　the　classification　loss　and　the　domain　adaptation　loss.　Model　comparison　experiments　and　ablation　experiments　were　designed,　and　the　F1　value　of　the　model　reached　79郾　61%,　indicating　that　the　adversarial　transfer　learning　model　has　good　performance　and　is　suitable　for　application　in　the　the　task　of　detecting　environmental　false　complaints　and　reports.　©　2023　Beijing　University　of　Technology.　All　rights　reserved.