摘要
近年,随着智能家居行业的火热,智能家居应用的开发也愈发重要。在智能家居应用智能化的软件需求下,该文旨在解决开发客户端语音控制功能时遇到的中文字符串识别精度问题。在智能家居场景下,语音控制指令字符串往往具有动词+关键词+家居名词的三段式结构,针对这一个特性,在对比分析了以往的中文字符串匹配算法之后,利用正向最大匹配算法进行分词,并结合汉字拼音码和汉字字形结构对编辑距离算法进行改进。最终提出了一种适合智能家居语音控制场景下的中文字符串匹配算法。研究结果表明该算法是有效,可行的。
In recent years,with the fiery development of the smart home industry,the development of smart home applications has become increasingly important. With the smart home application of intelligent software requirements,this paper aims to solve the problem of Chinese string matching accuracy encountered when developing client voice control functions. In the smart home scenario,the voice control instruction string often has a three-stage structure of the verb + key + home nouns. After comparing and analyzing the previous Chinese string matching algorithms,this paper uses the forward maximum matching algorithm to divide the word segmentation,and improve the editing distance algorithm combined with the Chinese character phonetic code and the Chinese character configuration. Finally,a Chinese string matching algorithm is proposed which is suitable for smart home voice control scenarios. The experimental results show that this algorithm is effective,feasible.
作者
吴振华
高瑞泽
WU Zhen-hua;GAO Rui-Ze(School of Software,Nanchang ttangkong University,Nanchang 330063,Chin)
出处
《南昌航空大学学报(自然科学版)》
CAS
2018年第2期81-85,共5页
Journal of Nanchang Hangkong University(Natural Sciences)
基金
国家自然科学基金(61762067)
关键词
智能家居
字符串匹配
分词
编辑距离
smarthome
string matching
word segmentation
editing distance