基于通用知识图谱的个性化智能助手  

Personalized intelligent assistant by integrating general knowledge graph

作　　者：林章颖 陈佳雯 陈星[1,2] LIN Zhangying;CHEN Jiawen;CHEN Xing(College of Computer and Data Science,Fuzhou University,Fuzhou Fujian 350108,China;Fujian Provincial Key Laboratory of Network Computing and Intelligent Information Processing(Fuzhou University),Fuzhou Fujian 350108,China)

机构地区：[1]福州大学计算机与大数据学院,福州350108 [2]福建省网络计算与智能信息处理重点实验室(福州大学),福州350108

出　　处：《计算机应用》2022年第S01期12-20,共9页journal of Computer Applications

基　　金：福建省自然科学基金杰青项目(2020J06014)。

摘　　要：针对现有个性化智能助手对预训练的语料数量和质量要求较高、无法根据用户的多变需求快速添加自然语言指令以触发智能助手的个性化场景等问题,提出一种基于通用知识图谱的个性化智能助手。首先,使用通用知识图谱的结构化的语义知识为用户的自然语言指令和应用场景进行语义标注并快速构建特征模型,弥合用户个性化的需求和对应场景的语义鸿沟;其次,使用余弦相似度计算用户指令与场景应用程序编程接口(API)的相似程度,进行API的匹配与推荐;最后,通过特征匹配对指令中API所需参数进行抽取并执行。实验结果表明,该方法指令匹配结果的宏F1值为84.72%,API参数抽取准确率达84.33%。Existing personalized intelligent assistants have high requirements on the quantity and quality of pre-trained corpus,and it is impossible to quickly add natural language instructions to trigger the personalized scene of the intelligent assistant according to the changing needs of users.In order to solve the problems above,a personalized intelligent assistant by integrating general knowledge graph was proposed.Firstly,natural language instructions and application scenarios of users were semantically labeled and he feature model was rapidly built tby the general knowledge graph,which bridged the semantic gap between user personalized needs and corresponding scenarios.After that,cosine similarity was used to calculate the similarity between the user instructions and the Application Programming Interfaces(API)of scenarios,by which to complete the API matching and recommendation.Finally,the parameters required by the API were extracted and the APIs were executed through feature matching.Experimental results show that the personalized assistant can provide an instruction matching result with the macro F1 of 84.72%,and the API parameter extraction accuracy of 84.33%.

关 键 词：智能助手 通用知识图谱 自然语言处理 相似度计算 语义理解 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]