轻量化人工智能翻译文本特征分类算法

Lightweight AI-translated text feature classification algorithm

由于人工智能翻译文本整体规模较大,在分类处理时往往存在领域划分异常的情况。为此,文章提出了轻量化人工智能翻译文本特征分类算法,构建了与特定领域相关的领域知识语料库,分别从词汇特征与句法特征2个角度提取人工智能翻译文本的轻量化特征。根据翻译文本特征与对应领域知识语料库特征之间的距离关系,该算法可实现分类处理,在对不同领域文本进行分类时不仅表现出较高的稳定性,且被准确分类文本数量始终保持在18篇以上,具有良好的分类效果。

Due to the large overall scale of text translation by artificial intelligence,there are often cases of domain segmentation anomalies during classification processing.To this end,the article proposes a lightweight artificial intelligence translation text feature classification algorithm,constructs domain knowledge corpora related to specific domains,extracts lightweight features of AI translated texts from two perspectives:lexical features and syntactic features.Based on the distance relationship between translated text features and corresponding domain knowledge corpus features,this algorithm can achieve classification processing.When classifying texts in different domains,it not only shows high stability,but also the number of accurately classified texts remains stable at 18 or more,with good classification performance.