基于DBNet和改进的Trie树搜索的网络敏感词检测技术

Network Sensitive Word Detection Technology Based on DBNet and Improved Trie Tree Search

为进一步避免未成年人接触到网络中的不良信息,提出一种基于DBNet和改进的Trie树搜索的网络敏感词检测方法。其中,以DBNet网络为基础的文本检测方法,以Trie树为基础的敏感词检测方法。实验结果表明,与其他文本检测方法以及文本识别方法相比,设计使用的文本检测和识别方法具有更高的精度,能够为后续的敏感词检测提供更加准确的文本信息;与传统的敏感词检测方法相比,基于DBNet和改进的Trie树搜索的敏感词检测方法具有更高的检测精度,检测准确率、漏检率以及误检率分别为89.12%、5.80%和6.12%。综上可知,设计的敏感词检测方法检测性能良好,精度较高,能够应用于实际的网络信息敏感词检测中,保护未成年人接触到网络中的不良信息,具有一定的可行性。

To further avoid minors from accessing harmful information on the internet,a network sensitive word detection method based on DBNet and improved Trie tree search is proposed.Among them,the text detection method based on DBNet network and the sensitive word detection method based on Trie tree.The experimental results show that compared with other text detection and recognition methods,the designed text detection and recognition method has higher accuracy and can provide more accurate text information for subsequent sensitive word detection;Compared with traditional sensitive word detection methods,the sensitive word detection method based on DBNet and improved Trie tree search has higher detection accuracy,with detection accuracy,missed detection rate,and false detection rate of 89.12%,5.80%,and 6.12%,respectively.In summary,it can be seen that the designed sensitive word detection method has good detection performance and high accuracy,and can be applied to practical network information sensitive word detection,protecting minors from accessing harmful information in the network,and has certain feasibility.