%0 Journal Article
%T 基于N-gram剪枝技术的隐患文本自动评估模型<br>An Automatic Assessment Model Based on N-gram Pruning Technique for Hidden Danger Text
%A 叶洪胜
%A 刘洪
%A 周宝山
%A 兰莉
%A 邹巧兰
%A 周啟梦
%A 王海宇
%J Mine Engineering
%P 388-394
%@ 2329-731X
%D 2024
%I Hans Publishing
%R 10.12677/me.2024.123047
%X 为了自动分析海上钻井平台隐患文本中蕴含的隐患响应程度信息，量化隐患严重程度，提出一种基于N-gram词袋向量的隐患响应等级量化评估模型。首先针对1565条钻井平台的现场隐患记录进行分词与过滤处理；其次再以N-gram作为特征单元重塑词袋维度；然后提出使用逆TF-IDF值来强化特征值；最后，使用朴素贝叶斯构建隐患量化模型。结果表明：使用该方法的隐患量化评估模型具有较高的精确率、召回率及F1值。<br />
To automatically analyze the response level information of hidden dangers contained in hidden danger texts and quantify the severity, a quantitative evaluation model based on N-gram word bag vectors is proposed for the response level of hidden dangers. Firstly, segment and filter the on-site hazard records of 1565 drilling platforms; Secondly, using N-gram as feature units to reshape the bag of words dimension; Then, it is proposed to use the inverse TF-IDF value to enhance the feature values; Finally, use naive Bayes to construct a hazard quantification model. The results show that the hazard quantification evaluation model using this method has high accuracy, recall, and F1 value.
%K 语义分析，钻井平台，N-gram，词袋向量，隐患量化<br>Semantic Analysis
%K Drilling Platforms
%K N-gram
%K Word Bag Vector
%K Hazard Quantification
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=91161