基于Rule-BertAtten的中文小说对话人物识别方法  

Rule-BertAtten Based Quote Attribution for Chinese Novels

作　　者：王子 王玉龙[1] 刘同存 李炜[1] 廖建新[1] WANG Zi;WANG Yulong;LIU Tongcun;LI Wei;LIAO Jianxin(State Key Laboratory of Networking and Switching Technology,Beijing University of Posts and Telecommunications,Beijing,100876,China;School of Information Engineering,Zhejiang A&F University,Hangzhou,Zhejiang 311300,China)

机构地区：[1]北京邮电大学网络与交换技术国际重点实验室,北京100876 [2]浙江农林大学信息工程学院,浙江杭州311300

出　　处：《中文信息学报》2022年第3期82-90,共9页Journal of Chinese Information Processing

基　　金：国家自然科学基金(61771068,61671079)。

摘　　要：小说中的对话人物识别任务是将小说中对话的说话者归属识别为小说中某个具体的人物,是有声小说自动合成的基础。为了能够充分表示对话类型的区别以及表示文本前后的语义特征,该文提出了一种基于Rule-BertAtten的中文小说对话人物识别方法。首先将对话主要分成四类,即有明确人物名作为主语的对话、人称代词性别唯一匹配候选人作为主语的对话、人称代词性别多匹配候选人作为主语的对话以及其他无任何特征作为主语的对话,根据对话的类别,采用规则判断和加入注意力机制的BERT词向量语义表示的方法,实验表明,该方法具有更高的准确率。Quote attribution in novels aims at determining who says a quote in a given novel.This task is important for assigning appropriate voices to the given quotes when producing vocal novels.In order to fully express the difference of quote types and the semantic features in the context,this paper proposes a Rule-BertAtten method for quote attribution in Chinese novels.The quotes are divided into four categories:the quote with explicit speaker,the quote with pronoun speaker with one-match gender,the quote with pronoun speaker with multi-match gender and the quote with implicit speaker.According to these categories,a rule-based method and the BERT word embedding methods with Attention are applied respectively.The experiment result shows that our method is more accurate than previous approaches.

关 键 词：小说对话 BERT 对话人物识别 规则法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]