WO2017016286A1 - 一种多语言语义解析的方法和装置 - Google Patents

一种多语言语义解析的方法和装置 Download PDF

Info

Publication number
WO2017016286A1
WO2017016286A1 PCT/CN2016/082653 CN2016082653W WO2017016286A1 WO 2017016286 A1 WO2017016286 A1 WO 2017016286A1 CN 2016082653 W CN2016082653 W CN 2016082653W WO 2017016286 A1 WO2017016286 A1 WO 2017016286A1
Authority
WO
WIPO (PCT)
Prior art keywords
semantic
control instruction
voice control
language
grammatical
Prior art date
Application number
PCT/CN2016/082653
Other languages
English (en)
French (fr)
Inventor
刘伟
Original Assignee
中兴通讯股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Publication of WO2017016286A1 publication Critical patent/WO2017016286A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00Handling natural language data
    • G06F40/30Semantic analysis
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00Handling natural language data
    • G06F40/20Natural language analysis
    • G06F40/205Parsing

Definitions

  • Embodiments of the present invention relate to, but are not limited to, an electronic terminal device, and in particular, to a multi-language semantic analysis method and apparatus.
  • the embodiment of the invention provides a multi-language semantic parsing method and device, which can make the semantic parsing simple, flexible and unlock the terminal.
  • a method for multilingual semantic parsing including:
  • the semantic analysis result corresponding to the voice control instruction is parsed according to the hierarchical structure of the grammatical structure tree.
  • the method further includes: performing an operation corresponding to the semantic parsing result.
  • the method before receiving the voice control instruction input by the user, the method further includes:
  • the grammatical structure identification of the voice control instruction according to the recognized language, and obtaining the grammatical structure tree with the semantic label includes:
  • the semantic tag includes one or more of the following: a domain, an action, an action identifier, and a keyword.
  • determining one or more semantic tags corresponding to the voice control instruction syntax includes:
  • a corresponding one or more semantic tags are determined for each word, phrase, and sentence obtained.
  • the embodiment of the invention further provides a device for multi-language semantic parsing, comprising:
  • a voice receiving module configured to receive a voice control instruction input by a user
  • a voice engine identification module configured to identify a language corresponding to the voice control instruction
  • a syntax tree processing module configured to perform grammatical structure recognition on the voice control instruction according to the recognized language, to obtain a grammatical structure tree with a semantic tag;
  • the semantic parsing processing module is configured to parse the semantic parsing result corresponding to the voice control instruction according to the hierarchical structure of the grammar structure tree.
  • the device further includes:
  • the control execution module is configured to perform an operation corresponding to the semantic analysis result.
  • the device further includes:
  • a grammar compilation module set to write one or more grammar files according to the language according to the Bacchus paradigm.
  • syntax tree processing module includes:
  • a grammatical component analysis unit configured to perform grammatical component analysis on the voice control instruction according to the level of the grammatical structure of the recognized language, to obtain a grammatical structure tree corresponding to the voice control instruction;
  • a semantic tag identifier unit configured to determine one or more semantic tags corresponding to the sentence structure of the voice control instruction
  • the grammar structure tree constituting unit is configured to combine the semantic tag and the grammar structure tree to obtain the grammar structure tree with the semantic tag.
  • the semantic tag identification unit is set to:
  • a corresponding one or more semantic tags are determined for each word, phrase, and sentence obtained.
  • Embodiments of the present invention further provide a computer readable storage medium storing computer executable instructions for performing any of the above multilingual semantic parsing methods.
  • the invention has the following beneficial effects:
  • the multi-language semantic parsing method and device provided by the invention can automatically recognize the corresponding language without considering the language problem, thereby realizing the use of no language barrier and increasing the convenience in practical use of the terminal.
  • FIG. 1 is a flow chart of a method for multi-language semantic parsing according to an embodiment of the present invention
  • FIG. 2 is a schematic structural diagram of an apparatus for multi-language semantic parsing according to an embodiment of the present invention.
  • an embodiment of the present invention provides a method for multi-language semantic parsing, including:
  • Step S101 Receive a voice control instruction input by a user, and identify a language corresponding to the voice control instruction;
  • Step S102 Perform grammatical structure recognition on the voice control instruction according to the recognized language, and obtain a grammatical structure tree with a semantic tag;
  • Step S103 Parse the semantic analysis result corresponding to the voice control instruction according to the hierarchical structure of the syntax structure tree.
  • the language corresponding to the voice control instruction is identified in step S101 to be identified according to different language model resources.
  • step S103 the method further includes:
  • Step S104 Perform an operation corresponding to the semantic analysis result.
  • the method further includes:
  • the Bacchus paradigm was first introduced by John Backus and Peter Naur with a formal notation to describe the grammar of a given language (the earliest used to describe the ALGOL 60 (algorithm language), the first generation of advanced in the history of computer development. Program language family) programming language).
  • the grammar files are written according to different languages, and the grammar files required to be written can express the domain of the voice control instruction, the action, the action ID and the keyword according to the level of the grammatical structure.
  • a subset of sentences (a set of one or more words, phrases, and sentences obtained through sentence structure analysis) identifies the relevant semantic tags of the sentence subset using TAG tags.
  • step S103 the domain of the voice control instruction, the action, and the action ID are parsed layer by layer according to the hierarchical structure of the syntax tree. If there are required keywords, such as "contact name”, "25 degrees” and other keywords are also parsed according to the corresponding semantic tags, and finally a set of related semantic analysis result instruction sets is determined.
  • Step S102 includes:
  • Step S1021 performing syntax component analysis on the voice control instruction according to the level of the recognized grammatical structure of the language, and obtaining a grammatical structure tree corresponding to the voice control instruction;
  • Step S1022 determining one or more semantic tags corresponding to the sentence structure of the voice control instruction
  • Step S1023 combines the semantic tag and the grammar structure tree to obtain the grammar structure tree with the semantic tag.
  • the semantic tag in step S1022 includes one or more of the following: a domain, an action, an action identifier, and a keyword.
  • Step S1022 includes:
  • a sentence structure of the voice control instruction is analyzed to obtain one or more words, phrases, and sentences; and one or more semantic tags are determined for each word, phrase, and sentence obtained.
  • an embodiment of the present invention further provides an apparatus for multi-language semantic parsing, including:
  • the voice receiving module 201 is configured to receive a voice control instruction input by the user
  • the voice engine identification module 202 is configured to identify a language corresponding to the voice control instruction
  • a syntax tree processing module 203 configured to perform grammatical structure recognition on the voice control instruction according to the recognized language, to obtain a grammatical structure tree with a semantic tag;
  • the semantic analysis processing module 204 is configured to parse the semantic analysis result corresponding to the voice control instruction according to the hierarchical structure of the syntax structure tree.
  • the device further includes:
  • the control execution module 205 is configured to perform an operation corresponding to the semantic analysis result.
  • the grammar compilation module 200 is configured to write one or more grammar files according to the language according to the Bacchus paradigm.
  • the syntax tree processing module 203 includes:
  • the grammatical component analysis unit 2031 is configured to perform grammatical component analysis on the voice control instruction according to the level of the grammatical structure of the recognized language, to obtain a grammatical structure tree corresponding to the voice control instruction;
  • a semantic tag identifying unit 2032 configured to determine one of the voice control instruction sentence patterns or Multiple semantic tags
  • the grammar structure tree structuring unit 2033 is configured to combine the semantic tag and the grammar structure tree to obtain the grammar structure tree with the semantic tag.
  • the semantic tag identification unit 2032 is used to:
  • a corresponding one or more semantic tags are determined for each word, phrase, and sentence obtained.
  • Speech recognition content returns a syntax tree structure with semantic tags according to the syntax structure written in the first step.
  • the possible forms are as follows:
  • the domain (Domain), action (Action), and action ID (Action ID) of the voice control instruction are parsed. If there are keywords that are needed, such as "contact name”, Keywords such as "25 degrees” are also parsed according to the corresponding semantic tags, and finally a set of related semantic parsing result instruction sets are output.
  • keywords that are needed such as "contact name”
  • Keywords such as "25 degrees” are also parsed according to the corresponding semantic tags, and finally a set of related semantic parsing result instruction sets are output.
  • the final result parsed by the semantic tag based on the [send short message] in this embodiment may be in the following form:
  • the final result of the present embodiment [adjusting the temperature of the air conditioner] after parsing based on the semantic tag may be in the following form:
  • the method and device for multi-language semantic parsing proposed by the embodiments of the present invention relate to the field of electronic terminal devices, and the method includes: receiving a voice control instruction input by a user, and identifying a language corresponding to the voice control instruction;
  • the language performs grammatical structure recognition on the voice control instruction to obtain a grammatical structure tree with a semantic tag; and parses out a semantic analysis result corresponding to the voice control instruction according to a hierarchical structure of the grammatical structure tree.
  • the language can be automatically recognized regardless of the language, to achieve the use of language-free obstacles, and to increase the convenience of the terminal in actual use.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Computational Linguistics (AREA)
  • General Health & Medical Sciences (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Machine Translation (AREA)

Abstract

一种多语言语义解析的方法和装置,涉及电子终端设备领域,所述方法包括:接收用户输入的语音控制指令,并识别所述语音控制指令对应的语言;按照识别出的所述语言对所述语音控制指令进行语法结构识别,获得带有语义标签的语法结构树;按照语法结构树的层级结构解析出所述语音控制指令对应的语义解析结果。可以不考虑语言的问题,自动识别对应的语言,来实现无语言障碍的使用,增加终端实际使用中的便利性。

Description

一种多语言语义解析的方法和装置 技术领域
本发明实施例涉及但不限于电子终端设备,尤指一种多语言语义解析的方法和装置。
背景技术
在语音控制指令领域,需要对所识别出的命令,比如:“给某某打电话”,“把空调的温度调到25度”等进行语义的解析,才能清楚地知道用户想要表达的操控意图,进而按照语音指令对目标终端进行相应控制。
但是,在应用的过程中,如果更换一种语言来表达,由于语法结构的完全不同,因此语义解析需要用两套甚至更多不同的逻辑来进行解析,这样,使得语义解析变得复杂、不灵活。
发明内容
以下是对本文详细描述的主题的概述。本概述并非是为了限制权利要求的保护范围。
本发明实施例提供一种多语言语义解析的方法和装置,能够使得语义解析变得简单、灵活,无障碍解锁终端。
一种多语言语义解析的方法,包括:
接收用户输入的语音控制指令,并识别语音控制指令对应的语言;
按照识别出的语言对语音控制指令进行语法结构识别,获得带有语义标签的语法结构树;
按照语法结构树的层级结构解析出语音控制指令对应的语义解析结果。
可选地,所述方法还包括:执行所述语义解析结果对应的操作。
可选地,所述接收用户输入的语音控制指令之前还包括:
按照巴科斯范式根据语言编写一种或者多种语法文件。
可选地,所述按照识别出的语言对语音控制指令进行语法结构识别,获得带有语义标签的语法结构树包括:
按照识别出的所述语言的语法结构的层级对所述语音控制指令进行语法成分分析,获得所述语音控制指令对应的语法结构树;
确定所述语音控制指令句式对应的一个或者多个语义标签;
结合所述语义标签和所述语法结构树,获得所述获得带有语义标签的语法结构树。
可选地所述语义标签包括以下的一项或者多项:领域、动作、动作标识、关键词。
可选地,确定所述语音控制指令句式对应的一个或者多个语义标签包括:
分析所述语音控制指令的句式结构,获得一个或者多个词语、短语和句子;
为获得的每个词语、短语和句子确定对应的一个或者多个语义标签。
本发明实施例还提供一种多语言语义解析的装置,包括:
语音接收模块,设置为接收用户输入的语音控制指令;
语音引擎识别模块,设置为识别所述语音控制指令对应的语言;
语法树处理模块,设置为按照识别出的所述语言对所述语音控制指令进行语法结构识别,获得带有语义标签的语法结构树;
语义解析处理模块,设置为按照语法结构树的层级结构解析出所述语音控制指令对应的语义解析结果。
可选地,所述装置还包括:
控制执行模块,设置为执行所述语义解析结果对应的操作。
可选地,所述装置还包括:
语法编译模块,设置为按照巴科斯范式根据语言编写一种或者多种语法文件。
可选地,所述语法树处理模块包括:
语法成分分析单元,设置为按照识别出的所述语言的语法结构的层级对所述语音控制指令进行语法成分分析,获得所述语音控制指令对应的语法结构树;
语义标签标识单元,设置为确定所述语音控制指令句式对应的一个或者多个语义标签;
语法结构树构成单元,设置为结合所述语义标签和所述语法结构树,获得所述获得带有语义标签的语法结构树。
可选地,所述语义标签标识单元设置为:
分析所述语音控制指令的句式结构,获得一个或者多个词语、短语和句子;
为获得的每个词语、短语和句子确定对应的一个或者多个语义标签。
本发明实施例再提供了一种计算机可读存储介质,存储有计算机可执行指令,所述计算机可执行指令用于执行上述任一多语言语义解析的方法。
本发明和现有技术相比,具有如下有益效果:
本发明提供的多语言语义解析的方法和装置,可以不考虑语言的问题,自动识别对应的语言,来实现无语言障碍的使用,增加终端实际使用中的便利性。
本发明实施例的其它特征和优点将在随后的说明书中阐述,并且,部分地从说明书中变得显而易见,或者通过实施本发明而了解。本发明的目的和其他优点可通过在说明书、权利要求书以及附图中所特别指出的结构来实现和获得。
在阅读并理解了附图和详细描述后,可以明白其他方面。
附图概述
此处所说明的附图用来提供对本发明实施例的进一步理解,构成本申请的一部分,本发明的示意性实施例及其说明用于解释本发明,并不构成对本发明的不当限定。在附图中:
图1是本发明实施例的一种多语言语义解析的方法的流程图;
图2是本发明实施例的一种多语言语义解析的装置的结构示意图。
本发明的较佳实施方式
为使本发明的目的、技术方案和优点更加清楚明白,下文中将结合附图对本发明的实施例进行详细说明。需要说明的是,在不冲突的情况下,本申请中的实施例及实施例中的特征可以相互任意组合。
如图1所示,本发明实施例提供一种多语言语义解析的方法,包括:
步骤S101、接收用户输入的语音控制指令,并识别所述语音控制指令对应的语言;
步骤S102、按照识别出的语言对所述语音控制指令进行语法结构识别,获得带有语义标签的语法结构树;
步骤S103、按照语法结构树的层级结构解析出所述语音控制指令对应的语义解析结果。
步骤S101中识别所述语音控制指令对应的语言是根据不同的语言模型资源进行识别。
步骤S103之后还包括:
步骤S104、执行所述语义解析结果对应的操作。
步骤S101之前还包括:
S100、按照巴科斯范式根据语言编写一种或者多种语法文件。
巴科斯范式是由John Backus和Peter Naur首次引入一种形式化符号来描述给定语言的语法(最早用于描述ALGOL 60(算法语言(ALGOrithmic Language)的缩写,是计算机发展史上首批产生的高级程式语言家族)编程语言)。按照巴科斯范式根据不同的语言编写语法文件,要求编写的语法文件能够按照语法结构的层级表现语音控制指令的领域(Domain),动作(Action),动作ID(Action ID)以及关键词,在每一个句式子集(经过句式结构分析获得一个或者多个词语、短语和句子组成的集合)用TAG标签标识出句式子集的相关语义标签。
步骤S103中按照语法树的层级结构一层一层地解析出语音控制指令的领域(Domain),动作(Action),动作ID(Action ID)。如果有需要的关键词,例如“联系人姓名”,“25度”等关键词也根据相对应的语义标签进行解析,最后确定一套相关的语义解析结果指令集。
步骤S102包括:
步骤S1021按照识别出的所述语言的语法结构的层级对所述语音控制指令进行语法成分分析,获得所述语音控制指令对应的语法结构树;
步骤S1022确定所述语音控制指令句式对应的一个或者多个语义标签;
步骤S1023结合所述语义标签和所述语法结构树,获得所述获得带有语义标签的语法结构树。
步骤S1022中的所述语义标签包括以下的一项或者多项:领域、动作、动作标识、关键词。
步骤S1022包括:
分析所述语音控制指令的句式结构,获得一个或者多个词语、短语和句子;为获得的每个词语、短语和句子确定对应的一个或者多个语义标签。
如图2所示,本发明实施例还提供一种多语言语义解析的装置,包括:
语音接收模块201,用于接收用户输入的语音控制指令;
语音引擎识别模块202,用于识别所述语音控制指令对应的语言;
语法树处理模块203,用于按照识别出的所述语言对所述语音控制指令进行语法结构识别,获得带有语义标签的语法结构树;
语义解析处理模块204,用于按照语法结构树的层级结构解析出所述语音控制指令对应的语义解析结果。
所述的装置还包括:
控制执行模块205,用于执行所述语义解析结果对应的操作。
语法编译模块200,用于按照巴科斯范式根据语言编写一种或者多种语法文件。
其中,语法树处理模块203包括:
语法成分分析单元2031,用于按照识别出的所述语言的语法结构的层级对所述语音控制指令进行语法成分分析,获得所述语音控制指令对应的语法结构树;
语义标签标识单元2032,用于确定所述语音控制指令句式对应的一个或 者多个语义标签;
语法结构树构成单元2033,用于结合所述语义标签和所述语法结构树,获得所述获得带有语义标签的语法结构树。
语义标签标识单元2032用于:
分析所述语音控制指令的句式结构,获得一个或者多个词语、短语和句子;
为获得的每个词语、短语和句子确定对应的一个或者多个语义标签。
实施例1
本实施例以以下例句作为示例:
【发送短信】
给张三发短信
吉田さんにメッセージを送る
Send sms to Mike
【调整空调的温度】
把空调的温度调到25度
エアコンを25度へちょうせいする
Turn the air conditioning temperature to 25degrees
1.根据不同的语言进行语法的编写,这里因为要展示区分领域Domain所以写了两类操控的例句,语法的详细展开只以发短信作为示例,调整空调的写法类似不再赘述,可能的形式如下:
Figure PCTCN2016082653-appb-000001
Figure PCTCN2016082653-appb-000002
Figure PCTCN2016082653-appb-000003
2.语音识别内容根据第一步所编写的语法结构返回带有语义标签的语法树结构,可能的形式如下:
Figure PCTCN2016082653-appb-000004
Figure PCTCN2016082653-appb-000005
Figure PCTCN2016082653-appb-000006
Figure PCTCN2016082653-appb-000007
3.按照语法树的层级结构解析出语音控制指令的领域(Domain),动作(Action),动作ID(Action ID)。如果有需要的关键词,例如“联系人姓名”, “25度”等关键词也根据相对应的语义标签进行解析,最后输出一套相关的语义解析结果指令集,可能的形式如下:
获取第一层的语义标签区分Domain,SENDSMS_TAG标志为发短信的领域Domain="send_sms",同样AIRCONCTRL_TAG标志为空调控制的领域Domain="aircon_ctrl"。获取第二层的语义标签区分动作,发短信可能会分为单命令发短信Action="send_sms_only",发短信给联系人Action="send_sms_name"等动作。动作下面没有需要区分的ID,因此本例没有Action ID的解析。在发短信给某某这个示例中是需要解析联系人这个关键字的,因为手机终端在调起发短信这个操作的同时是需要知道联系人是谁才能做出正确的操作的。因此可能的解析为找到name这个语义标签,找到关键字,这个地方可能会是Arg1="张三"。
因此本实施例【发送短信】经过基于语义标签解析出的最终结果可能是以下形式:
中文:
{"domain":"send_sms","action":"send_sms_name","name":"张三"}
日文:
{"domain":"send_sms","action":"send_sms_name","name":"吉田"}
英文:
{"domain":"send_sms","action":"send_sms_name","name":"Mike"}
本实施例【调整空调的温度】经过基于语义标签解析出的最终结果可能是以下形式:
中文,日文,英文:
{"domain":"aircon_ctrl","action":"change_temp","temperature":"25"}
因为动作可能会有区分的原因,这里再举一个控制空调开关的例子“关闭空调”和“打开空调”,对于该例可能分别会是以下形式,action=0表示关,action=1表示开:
{"domain":"aircon_ctrl","action":"switch","action_id":"0"}
{"domain":"aircon_ctrl","action":"switch","action_id":"1"}
4.按照所输出的语义解析结果指令集进行对设备终端的实际操控。由上一步可以看出该指令集已经是跟语言无关,终端设备可以直接根据该指令集做出语音指令想要的相关操作,例如{"domain":"send_sms","action":"send_sms_name","name":"张三"}
由domain和action,可以知道是需要调起短信模块发短信界面,name可以知道是需要将张三填到联系人栏中,同样语言不同,仅仅是联系人名填得不同,跟语言是无关的。
以上所述仅为本发明的优选实施例而已,并不用于限制本发明,对于本领域的技术人员来说,本发明可以有各种更改和变化。凡在本发明的精神和原则之内,所作的任何修改、等同替换、改进等,均应包含在本发明的保护范围之内。本领域普通技术人员可以理解上述方法中的全部或部分步骤可通过程序来指令相关硬件完成,所述程序可以存储于计算机可读存储介质中,如只读存储器、磁盘或光盘等。可选地,上述实施例的全部或部分步骤也可以使用一个或多个集成电路来实现,相应地,上述实施例中的各模块/模块可以采用硬件的形式实现,也可以采用软件功能模块的形式实现。本申请不限制于任何特定形式的硬件和软件的结合。
以上仅为本发明的优选实施例,并非因此限制本发明的专利范围,凡是利用本发明说明书及附图内容所作的等效结构或等效流程变换,或直接或间接运用在其他相关的技术领域,均同理包括在本发明的专利保护范围内。
工业实用性
本发明实施例提出的多语言语义解析的方法和装置,涉及电子终端设备领域,方法,所述方法包括:接收用户输入的语音控制指令,并识别所述语音控制指令对应的语言;按照识别出的所述语言对所述语音控制指令进行语法结构识别,获得带有语义标签的语法结构树;按照语法结构树的层级结构解析出所述语音控制指令对应的语义解析结果。可以不考虑语言的问题,自动识别对应的语言,来实现无语言障碍的使用,增加终端实际使用中的便利性。

Claims (12)

  1. 一种多语言语义解析的方法,其特征在于,包括:
    接收用户输入的语音控制指令,并识别语音控制指令对应的语言;
    按照识别出的语言对语音控制指令进行语法结构识别,获得带有语义标签的语法结构树;
    按照语法结构树的层级结构解析出语音控制指令对应的语义解析结果。
  2. 如权利要求1所述的方法,该方法还包括:执行所述语义解析结果对应的操作。
  3. 如权利要求1所述的方法,所述接收用户输入的语音控制指令之前还包括:
    按照巴科斯范式根据语言编写一种或者多种语法文件。
  4. 如权利要求1所述的方法,其中,所述按照识别出的语言对语音控制指令进行语法结构识别,获得带有语义标签的语法结构树包括:
    按照识别出的所述语言的语法结构的层级对所述语音控制指令进行语法成分分析,获得所述语音控制指令对应的语法结构树;
    确定所述语音控制指令句式对应的一个或者多个语义标签;
    结合所述语义标签和所述语法结构树,获得所述获得带有语义标签的语法结构树。
  5. 如权利要求4所述的方法,其中,所述语义标签包括以下的一项或者多项:领域、动作、动作标识、关键词。
  6. 如权利要求4所述的方法,其中,所述确定所述语音控制指令句式对应的一个或者多个语义标签包括:
    分析所述语音控制指令的句式结构,获得一个或者多个词语、短语和句子;
    为获得的每个词语、短语和句子确定对应的一个或者多个语义标签。
  7. 一种多语言语义解析的装置,其特征在于,包括:
    语音接收模块,设置为接收用户输入的语音控制指令;
    语音引擎识别模块,设置为识别所述语音控制指令对应的语言;
    语法树处理模块,设置为按照识别出的所述语言对所述语音控制指令进行语法结构识别,获得带有语义标签的语法结构树;
    语义解析处理模块,设置为按照语法结构树的层级结构解析出所述语音控制指令对应的语义解析结果。
  8. 如权利要求7所述的装置,还包括:
    控制执行模块,设置为执行所述语义解析结果对应的操作。
  9. 如权利要求7所述的装置,还包括:
    语法编译模块,设置为按照巴科斯范式根据语言编写一种或者多种语法文件。
  10. 如权利要求7所述的装置,其中,所述语法树处理模块包括:
    语法成分分析单元,设置为按照识别出的所述语言的语法结构的层级对所述语音控制指令进行语法成分分析,获得所述语音控制指令对应的语法结构树;
    语义标签标识单元,设置为确定所述语音控制指令句式对应的一个或者多个语义标签;
    语法结构树构成单元,用于结合所述语义标签和所述语法结构树,获得所述获得带有语义标签的语法结构树。
  11. 如权利要求10所述的装置,其中,所述语义标签标识单元用于:
    分析所述语音控制指令的句式结构,获得一个或者多个词语、短语和句子;
    为获得的每个词语、短语和句子确定对应的一个或者多个语义标签。
  12. 一种计算机可读存储介质,存储有计算机可执行指令,所述计算机可执行指令用于执行权1~权6任一项的多语言语义解析的方法。
PCT/CN2016/082653 2015-07-24 2016-05-19 一种多语言语义解析的方法和装置 WO2017016286A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201510443910.1A CN106372054B (zh) 2015-07-24 2015-07-24 一种多语言语义解析的方法和装置
CN201510443910.1 2015-07-24

Publications (1)

Publication Number Publication Date
WO2017016286A1 true WO2017016286A1 (zh) 2017-02-02

Family

ID=57880209

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/082653 WO2017016286A1 (zh) 2015-07-24 2016-05-19 一种多语言语义解析的方法和装置

Country Status (2)

Country Link
CN (1) CN106372054B (zh)
WO (1) WO2017016286A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111143330A (zh) * 2019-12-27 2020-05-12 湖南亚信软件有限公司 一种多模态数据库解析引擎的实现方法及装置
CN112949286A (zh) * 2021-03-09 2021-06-11 北京汉雅天诚教育科技有限公司 一种基于句式结构的汉语自动句法分析器

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108399919A (zh) * 2017-02-06 2018-08-14 中兴通讯股份有限公司 一种语义识别方法和装置
CN109841210B (zh) * 2017-11-27 2024-02-20 西安中兴新软件有限责任公司 一种智能操控实现方法及装置、计算机可读存储介质
CN108228191B (zh) * 2018-02-06 2022-01-25 威盛电子股份有限公司 语法编译系统以及语法编译方法
CN110096709A (zh) * 2019-05-07 2019-08-06 百度在线网络技术(北京)有限公司 指令处理方法及装置、服务器及计算机可读介质
CN110970028B (zh) * 2019-12-26 2022-07-22 杭州中科先进技术研究院有限公司 一种规范语音识别设备的语音识别指令与操作指令的方法
CN111933141A (zh) * 2020-08-31 2020-11-13 江西台德智慧科技有限公司 一种基于大数据的人工智能语音交互系统
CN116955649A (zh) * 2023-07-21 2023-10-27 重庆赛力斯新能源汽车设计院有限公司 意图识别方法、装置、电子设备及存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1526132A (zh) * 2000-10-10 2004-09-01 ض� 基于语音的不受语言限制的搜索系统
US20050055209A1 (en) * 2003-09-05 2005-03-10 Epstein Mark E. Semantic language modeling and confidence measurement
CN101110812A (zh) * 2007-08-29 2008-01-23 中兴通讯股份有限公司 文本命令解析处理方法
CN104485106A (zh) * 2014-12-08 2015-04-01 畅捷通信息技术股份有限公司 语音识别方法、语音识别系统和语音识别设备

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0869466A (ja) * 1994-08-30 1996-03-12 Sumitomo Electric Ind Ltd 自然言語解析装置
US7593843B2 (en) * 2004-03-30 2009-09-22 Microsoft Corporation Statistical language model for logical form using transfer mappings

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1526132A (zh) * 2000-10-10 2004-09-01 ض� 基于语音的不受语言限制的搜索系统
US20050055209A1 (en) * 2003-09-05 2005-03-10 Epstein Mark E. Semantic language modeling and confidence measurement
CN101110812A (zh) * 2007-08-29 2008-01-23 中兴通讯股份有限公司 文本命令解析处理方法
CN104485106A (zh) * 2014-12-08 2015-04-01 畅捷通信息技术股份有限公司 语音识别方法、语音识别系统和语音识别设备

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111143330A (zh) * 2019-12-27 2020-05-12 湖南亚信软件有限公司 一种多模态数据库解析引擎的实现方法及装置
CN111143330B (zh) * 2019-12-27 2023-10-03 湖南亚信软件有限公司 一种多模态数据库解析引擎的实现方法及装置
CN112949286A (zh) * 2021-03-09 2021-06-11 北京汉雅天诚教育科技有限公司 一种基于句式结构的汉语自动句法分析器
CN112949286B (zh) * 2021-03-09 2023-08-04 北京汉雅天诚教育科技有限公司 一种基于句式结构的汉语自动句法分析器

Also Published As

Publication number Publication date
CN106372054A (zh) 2017-02-01
CN106372054B (zh) 2020-10-09

Similar Documents

Publication Publication Date Title
WO2017016286A1 (zh) 一种多语言语义解析的方法和装置
US10339916B2 (en) Generation and application of universal hypothesis ranking model
JP6909832B2 (ja) オーディオにおける重要語句を認識するための方法、装置、機器及び媒体
US9805718B2 (en) Clarifying natural language input using targeted questions
US20180061408A1 (en) Using paraphrase in accepting utterances in an automated assistant
US20180341698A1 (en) Method and apparatus for parsing query based on artificial intelligence, and storage medium
CN108369580B (zh) 针对屏幕上项目选择的基于语言和域独立模型的方法
US10242670B2 (en) Syntactic re-ranking of potential transcriptions during automatic speech recognition
CN109616096B (zh) 多语种语音解码图的构建方法、装置、服务器和介质
US9262411B2 (en) Socially derived translation profiles to enhance translation quality of social content using a machine translation
US11164562B2 (en) Entity-level clarification in conversation services
US11216292B2 (en) Adapting client application of feature phone based on experiment parameters
WO2018153273A1 (zh) 语义解析方法、装置及存储介质
US10754885B2 (en) System and method for visually searching and debugging conversational agents of electronic devices
US9400633B2 (en) Methods and apparatus for voiced-enabling a web application
WO2017166631A1 (zh) 语音信号处理方法、装置和电子设备
KR20190000776A (ko) 정보 입력 방법
US10885281B2 (en) Natural language document summarization using hyperbolic embeddings
US10223349B2 (en) Inducing and applying a subject-targeted context free grammar
US20140040745A1 (en) Methods and apparatus for voiced-enabling a web application
KR102102287B1 (ko) 챗봇을 위한 대화 모델의 데이터 크라우드소싱 방법
US10714087B2 (en) Speech control for complex commands
WO2019148797A1 (zh) 自然语言处理方法、装置、计算机设备和存储介质
KR20150045177A (ko) 사용자 발화 스타일에 따른 대화형 서비스 장치 및 방법
CN111209746A (zh) 自然语言处理方法、装置、存储介质及电子设备

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16829653

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 16829653

Country of ref document: EP

Kind code of ref document: A1