基于Chinese-BERT和BiLSTM的海洋石油工程行業(yè)質(zhì)量問題分類

打開文本圖片集

中圖分類號(hào)：TP3191.1；TP18 文獻(xiàn)標(biāo)識(shí)碼：A 文章編號(hào)：2096-4706（2025）07-0151-06

Abstract： Inorder toimprove theintelligencelevelofengineering construction quality inspection inofshoreoilindustry a quality problem clasification algorithm basedon Chinese pre-training modelof Chinese-BERTandBiLSTMis proposed. Chinese-BERTmodelcanbeter extract Chinese textfeaturesandimprovetheaccuracyoftextvectorsemanticrepresentationby using Chinese pre-training，whileBiLSTMfurtherextractskey informationfrom Chinese-BERTfeature vectors.Thevalidityof the modelis verified inrealdataandtheresults showthat the proposed modeloutperformsothercomparativemodels interms of classification performance，and it can effectively solve quality problem classfication problems.

eywords：offshore oil industry;quality problem; text classification; Chinese-BERT; BiLSTN

0 引言

海洋石油工程行業(yè)中，工程施工質(zhì)量檢查具有至關(guān)重要的地位和作用，通過施工質(zhì)量檢查獲取的質(zhì)檢數(shù)據(jù)是衡量與判斷工程項(xiàng)目質(zhì)量的重要標(biāo)準(zhǔn)和技術(shù)依據(jù)，高效的項(xiàng)目質(zhì)量檢測(cè)不僅能夠提高項(xiàng)目的施工質(zhì)量以及可維護(hù)性。（剩余8579字）