服务机器人自然语言处理的研究与应用  

Research and Application of Natural Language Processing in Service Robot

作　　者：王凡 尹浩伟 蒋峰岭 郭玉堂 WANG Fan;YIN Haowei;JIANG Fengling;GUO Yutang(College of Computer Science,Hefei Normal University,Hefei 230601,China)

机构地区：[1]合肥师范学院计算机学院,安徽合肥230601

出　　处：《安徽科技学院学报》2018年第4期61-69,共9页Journal of Anhui Science and Technology University

基　　金：安徽省高等学校自然科学研究重点项目(2013A217);合肥师范学院青年基金项目(2015QN15;2017QN19)

摘　　要：目的:以服务机器人的自然语言处理为目标,主要研究场景信息分析、自然语言任务信息理解等问题,最终求解出机器人的最优动作序列。方法:使用关键字搜索算法处理自然语言,通过格式化处理、分组标记和谓词转换等技术处理为特定格式的指令语言,之后使用回答集编程(Answer Set Programming,ASP)技术对指令语言任务信息进行规划,求解出最佳的机器人动作序列,并改进ASP在自然语言任务规划中的性能。结果:通过试验验证关键词搜索算法,在自然语言任务处理中较简单处理方式,有明显的性能提高。结论:关键字搜索算法能很好的完成自然语言处理任务,并可提高任务规划的性能,以及提升服务机器人自然语言处理的整体效率。Objective:Aiming at the natural language processing of service robot,the scene information and understanding of natural language task information were analyzed.It solves finally the optimal action sequence of robot.Methods:This article uses the keywords search algorithm to process natural language,and converts the natural to instruction language by using some technologies such as format process,group mark and conversion predicate.After that,the Answer Set Programming(ASP)technology is used to plan the task information to solve the best sequence of robot action and improve the performance of ASP in natural language task planning.Results:Experiments show that keyword search algorithm has obvious performance improvement in natural language task processing.Conclusion:The keywords search algorithm can complete the natural language processing task well,improve the performance of the task planning,and improve the overall efficiency of the natural language processing of service robot.

关 键 词：服务机器人 自然语言处理 关键字搜索 回答集编程 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]