WO2013174092A1 - 一种基于语音识别的查询方法及装置 - Google Patents

一种基于语音识别的查询方法及装置 Download PDF

Info

Publication number
WO2013174092A1
WO2013174092A1 PCT/CN2012/082965 CN2012082965W WO2013174092A1 WO 2013174092 A1 WO2013174092 A1 WO 2013174092A1 CN 2012082965 W CN2012082965 W CN 2012082965W WO 2013174092 A1 WO2013174092 A1 WO 2013174092A1
Authority
WO
WIPO (PCT)
Prior art keywords
voice
agent
customer
keyword
standard
Prior art date
Application number
PCT/CN2012/082965
Other languages
English (en)
French (fr)
Inventor
赵业
周洪凯
Original Assignee
华为技术有限公司
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 华为技术有限公司 filed Critical 华为技术有限公司
Priority to US13/899,110 priority Critical patent/US20130315385A1/en
Publication of WO2013174092A1 publication Critical patent/WO2013174092A1/zh

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04M—TELEPHONIC COMMUNICATION
    • H04M3/00—Automatic or semi-automatic exchanges
    • H04M3/42—Systems providing special services or facilities to subscribers
    • H04M3/487—Arrangements for providing information services, e.g. recorded voice services or time announcements
    • H04M3/493—Interactive information services, e.g. directory enquiries ; Arrangements therefor, e.g. interactive voice response [IVR] systems or voice portals
    • H04M3/4931—Directory assistance systems
    • H04M3/4933—Directory assistance systems with operator assistance
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04M—TELEPHONIC COMMUNICATION
    • H04M3/00—Automatic or semi-automatic exchanges
    • H04M3/42—Systems providing special services or facilities to subscribers
    • H04M3/487—Arrangements for providing information services, e.g. recorded voice services or time announcements
    • H04M3/493—Interactive information services, e.g. directory enquiries ; Arrangements therefor, e.g. interactive voice response [IVR] systems or voice portals
    • H04M3/4936—Speech interaction details
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04M—TELEPHONIC COMMUNICATION
    • H04M2201/00—Electronic components, circuits, software, systems or apparatus used in telephone systems
    • H04M2201/40—Electronic components, circuits, software, systems or apparatus used in telephone systems using speech recognition

Definitions

  • the present invention relates to the field of communications, and in particular, to a voice recognition based query method and apparatus.
  • the customer dials the access code and communicates with the agent through the CC (Call Center) system in the customer service system.
  • the CC system is generally integrated in the CRM (Customer Relationship Management) system.
  • the agent provides remote service to the customer through the CC system.
  • Commonly available are full manual services and speech recognition services.
  • the full manual service is provided by the agent to switch multiple service modules in the CRM software to find a related page of the customer query problem, and provides services for the customer; the voice recognition service performs problem location and automatic search through the voice recognition of the customer.
  • Related pages, and then agents provide services to customers.
  • the inventor finds that at least the following problems exist in the prior art:
  • the agent needs to switch multiple service modules to find related pages, and the search efficiency is low.
  • the application of the voice service may result in inaccurate customer voices due to non-standard pronunciation or non-standard terms, which may result in the inability of the customer voice to be accurately identified and required to be transferred from the voice recognition service to the full manual service.
  • Embodiments of the present invention provide a voice recognition-based query method and apparatus, which can be Improve the accuracy of speech recognition and improve work efficiency.
  • the embodiment of the present invention adopts the following technical solutions:
  • a voice recognition based query method including:
  • the query result is obtained according to the keyword of the standard voice of the agent, so that the agent provides services to the client according to the query result.
  • a voice recognition based query device comprising:
  • a determining unit configured to determine whether the received customer voice is a standard voice
  • a sending unit configured to: when the determining unit determines that the customer voice is a non-standard voice, forward the customer voice to an agent to prompt the agent to perform a standard pronunciation; and receive a unit, configured to receive the agent Standard voice
  • a keyword extracting unit configured to extract a keyword of the standard voice of the agent person received by the receiving unit
  • a query unit configured to perform, according to a keyword of the standard voice of the agent person extracted by the keyword extracting unit The query obtains the query result, so that the agent personnel provides services to the client according to the query result.
  • the voice recognition-based query method and device when determining that the received customer voice is a non-standard voice, forwarding the customer voice to an agent, to prompt the agent to perform a standard pronunciation; then extracting the received keyword of the standard voice of the agent; and querying according to the keyword of the standard voice of the agent to obtain a query result, so that the agent is based on the query result
  • the customer provides the service.
  • accurate speech recognition can be performed when the customer's pronunciation is not standard or the term is not standard, which improves the accuracy of speech recognition, accelerates the query rate, and improves work efficiency.
  • FIG. 1 is a schematic flowchart of a voice recognition based query method according to an embodiment of the present invention
  • FIG. 2 is a schematic flow chart of another voice recognition based query method according to an embodiment of the present invention.
  • FIG. 3 is a structural block diagram of a voice recognition based query device according to an embodiment of the present invention.
  • Embodiment 1 The embodiment of the present invention provides a query method based on voice recognition. The method shown in FIG. 1 includes the following steps:
  • the customer dials the access code to start accessing the customer service system, and sends the customer voice to the customer service system through the client terminal, and the customer describes the service to be performed by the customer in the customer voice.
  • the customer service system After receiving the customer voice, the customer service system applies a voice recognition technology to determine whether the customer voice is a standard voice.
  • the standard voice described herein includes both the pronunciation standard, that is, the standard Mandarin spoken by the customer, and the standard of the term, that is, the business terminology is required by the customer.
  • the customer service system determines that there is a standard business term in the customer voice, it determines that the received customer voice is a standard voice; otherwise, it determines that the received client voice is a non-standard voice.
  • the customer service system forwards the customer voice to the agent, and the agent hears the voice of the client and performs a standard pronunciation according to the non-standard voice.
  • the customer service system can receive the standard voice of the agent, and then use the voice to identify the keyword in the standard voice of the agent.
  • the customer service system After extracting the keywords in the standard voice of the agent, the customer service system automatically queries all the services according to the keyword, and obtains the query result related to the keyword.
  • the query results include all business links associated with the keyword, or business pages associated with the keyword.
  • the agent may select a service link required by the client in the service link, and click to directly enter the service page to provide services for the client;
  • the query result is a service page related to the keyword, and the agent can directly provide services to the client according to the service page.
  • the query rate is improved compared with the full manual service in the prior art, thereby improving the work efficiency of the agent;
  • the voice recognition service the voice is enhanced by the combination of the customer voice recognition and the voice recognition of the work and the voice.
  • the accuracy of the identification which in turn increases the query rate.
  • the method described in the foregoing steps 101 to 104 is an inquiry method performed when the client voice is not standard. Further, as shown in FIG. 2, the method further includes the following steps:
  • the customer service system determines that there is a standard business term in the customer voice, it determines that the received customer voice is a standard voice; otherwise, it determines that the received client voice is a non-standard voice.
  • the customer voice is a standard voice
  • extract keywords in the customer voice If the customer service system recognizes that there is a standard business term in the customer voice and determines that the customer voice is a standard voice, the customer service system directly extracts the keyword in the customer voice through the voice recognition technology.
  • 202. Perform a query according to a keyword in the customer voice to obtain the query result, so that the agent personnel provides a service for the client according to the query result.
  • the customer service system After extracting the keyword in the customer voice, the customer service system automatically queries all the services according to the keyword, and obtains a query result related to the keyword.
  • the query result includes all business links associated with the keyword, or a business page associated with the keyword.
  • the agent may select a service link required by the client in the service link, and click to directly enter the service page to provide services for the client;
  • the query result is a service page related to the keyword, and the agent can directly provide services to the client according to the service page. This improves the query rate compared to the full manual service in the prior art, thereby improving the work efficiency of the agent.
  • Lao Li's voice is standard voice.
  • Lao Li dialed the phone and said "I want to check the account information”.
  • the customer service system can identify the keyword "account”, and the system can automatically query to obtain a menu composed of all the business links related to the "account”.
  • Xiao Zhang can ask the specific business that Lao Li needs to carry out and select the corresponding business link in the menu, and directly enter the business page, so that the system automatically queries, and the query rate is 4 ⁇ .
  • Lao Li is a non-standard voice.
  • Lao Li dialed the general dialect and said, "I want to check the account” or "I want to check the account” in Mandarin the customer service system can not identify the standard business terms in these two pronunciations, so the voice of Lao Li is judged to be non-standard voice.
  • the customer service system will send "I want to check the account” to Xiao Zhang, Xiao Zhang will carry out the standard pronunciation "OK, I will help you to check the account profile”;
  • the system can identify the keyword “account” according to Xiao Zhang's pronunciation. " , through the system automatically query to get a menu of all the business links related to the "account”, at this time Xiao Zhang can ask the specific business that Lao Li needs to carry out and select the corresponding business link in the menu, directly enter the business page , this is automatically queried by the system, the query rate is 4 ⁇ high.
  • voice recognition can also be applied for voice recognition query.
  • the embodiment of the present invention further provides a voice recognition-based query device.
  • the device includes: a determining unit 301, a sending unit 302, a receiving unit 303, and a keyword
  • the unit 304 is used to query the unit 305.
  • the determining unit 301 is configured to determine whether the received customer voice is a standard voice.
  • the customer dials the access code to start accessing the customer service system, and sends the customer voice to the customer service system through the client terminal, and the customer describes the service to be performed by the customer in the customer voice.
  • the customer service system After receiving the customer voice, the customer service system applies a voice recognition technology to determine whether the customer voice is a standard voice.
  • the standard voice described herein includes both the pronunciation standard, that is, the standard Mandarin spoken by the customer, and the standard of the term, that is, the business terminology is required by the customer.
  • the determining unit 301 recognizes that there is a standard business term in the customer voice, it judges that the received customer voice is a standard voice; otherwise, it determines that the received client voice is a non-standard voice.
  • the sending unit 302 is configured to: when the determining unit 301 determines that the customer voice is a non-standard voice, forward the customer voice to an agent to prompt the agent to perform standard pronunciation.
  • the sending unit 302 forwards the customer voice to the agent, and the agent hears the voice of the client according to the non-standard voice. Make a standard pronunciation.
  • the receiving unit 303 is configured to receive a standard voice of the agent.
  • the keyword extracting unit 304 is configured to extract a keyword of the standard voice of the agent received by the receiving unit.
  • the receiving unit 303 can receive the standard voice of the agent, and then the keyword extracting unit 304 extracts the key words in the standard voice of the agent by the voice recognition technology.
  • the query unit 305 is configured to query the keyword according to the standard voice of the agent extracted by the keyword extracting unit to obtain a query result, so that the agent provides services for the client according to the query result.
  • the query unit 305 After the keyword extracting unit 304 extracts the keyword in the standard voice of the agent, the query unit 305 automatically queries all the services according to the keyword, and obtains a query result related to the keyword.
  • the query result includes all business links associated with the keyword, or a business page associated with the keyword.
  • the agent person may select a service link required by the client in the service link, and click directly to enter the service link.
  • the service page provides services for the client; if the query result is a service page related to the keyword, the agent can directly provide services to the client according to the service page.
  • the query rate is improved compared with the full manual service in the prior art, thereby improving the work efficiency of the agent; compared with the voice recognition service, the voice is enhanced by the combination of the customer voice recognition and the voice recognition of the work and the voice.
  • the accuracy of the identification which in turn increases the query rate.
  • the keyword extracting unit 304 is further configured to extract keywords in the customer voice when the client voice is a standard voice.
  • the customer service system directly extracts keywords in the customer voice through voice recognition technology.
  • the query unit 305 is further configured to perform the query according to the keyword in the customer voice to obtain the query result, so that the agent personnel provide services for the client according to the query result.
  • the query unit 305 After the keyword extracting unit 304 extracts the keyword in the customer voice, the query unit 305 automatically queries all the services according to the keyword, and obtains a query result related to the keyword.
  • the query result includes all business links associated with the keyword, or a business page associated with the keyword.
  • the agent may select a service link required by the client in the service link, and click to directly enter the service page to provide services for the client;
  • the query result is a service page related to the keyword, and the agent can directly provide services to the client according to the service page. This improves the query rate compared with the full manual service in the prior art, thereby improving the work efficiency of the agent.
  • the voice recognition-based query method and device when determining that the received customer voice is a non-standard voice, forwarding the customer voice to an agent, to prompt the agent to perform Standard voice; then extracting the received keyword of the standard voice of the agent; and querying according to the keyword of the standard voice of the agent to obtain the query result, so that the agent is based on the query result
  • the customer provides the service. In this way, if the pronunciation of the customer is not standard or the standard is not standard, accurate speech recognition can be performed, the accuracy of the speech recognition is improved, the query rate is accelerated, and the work efficiency is improved.

Landscapes

  • Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Human Computer Interaction (AREA)
  • Telephonic Communication Services (AREA)

Abstract

本发明实施例提供了一种基于语音识别的查询方法及装置，涉及通信领域，可以提升语音识别的准确性，提高了工作效率。所述方法包括：在判断接收到的所述客户语音为非标准语音时，将所述客户语音转发给座席人员，以提示所述座席人员进行标准发音；然后提取接收到的所述座席人员的标准语音的关键字；并根据所述座席人员的标准语音的关键字进行查询获得查询结果，以使得所述座席人员根据所述查询结果为所述客户提供服务。本发明实施例应用于客服系统。

Description

一种基于语音识别的查询方法及装置 本申请要求于 201 1 年 05 月 23 日提交中国专利局、 申请号为 201210161738.7、 发明名称为 "一种基于语音识别的查询方法及装 置" 的中国专利申请的优先权， 其全部内容通过引用结合在本申请 中。
技术领域
本发明涉及通信领域， 尤其涉及一种基于语音识别的查询方法 及装置。
背景技术
随着服务行业竟争日益激烈， 各服务行业对客户的满意度和意见反 馈越来越重视， 客服系统的建设越来越完善。 客户通过客服系统可以进 行查询花费、 了解业务情况、 反馈意见、 投诉、 订购或退订业务等多项 业务。
客户拨打接入码通过客服系统中的 CC ( Call Center, 呼叫中心 ) 系 统与座席人员进行交流， 所述 CC 系统一般集成在 CRM ( Customer Relationship Management, 客户关系管理） 系统中。 所述座席人员通过所 述 CC系统为客户提供远程服务。常见的有全人工服务和语音识别服务两 种。所述全人工服务为由所述座席人员在 CRM软件中切换多个业务模块 查找客户查询问题的相关页面， 为客户提供服务； 所述语音识别服务为 通过客户的语音识别进行问题定位， 自动查找相关页面， 再由座席人员 为客户提供服务。
在实现上述访问客服系统的过程中， 发明人发现现有技术中至少存 在如下问题： 在应用所述全人工服务时， 所述座席人员需要切换多个业 务模块才能查找到相关页面， 查找效率低下； 而应用所述语音服务时可 能会因为客户发音不标准或用语不标准， 导致客户语音无法准确识别， 需要由语音识别服务转至全人工服务， 导致效率低下。
发明内容 本发明的实施例提供一种基于语音识别的查询方法及装置， 可以提 升语音识别的准确性， 提高了工作效率。 为达到上述目的， 本发明的实施例采用如下技术方案：
一种基于语音识别的查询方法， 包括：
判断接收到的客户语音是否为标准语音；
若所述客户语音为非标准语音， 则将所述客户语音转发给座席人 员， 以提示所述座席人员进行标准发音；
接收所述座席人员的标准语音并提取所述座席人员的标准语音的关 键字；
根据所述座席人员的标准语音的关键字进行查询获得查询结果， 以 使得所述座席人员根据所述查询结果为所述客户提供服务。
一种基于语音识别的查询装置， 包括：
判断单元， 用于判断接收到的客户语音是否为标准语音；
发送单元， 用于在判断单元判断所述客户语音为非标准语音时， 将 所述客户语音转发给座席人员， 以提示所述座席人员进行标准发音； 接收单元， 用于接收所述座席人员的标准语音；
关键字提取单元， 用于提取所述接收单元接收到的所述座席人员的 标准语音的关键字； 查询单元， 用于根据关键字提取单元提取到的所述座席人员的标准 语音的关键字进行查询获得查询结果， 以使得所述座席人员根据所述查 询结果为所述客户提供服务。
本发明实施例提供的一种基于语音识别的查询方法及装置， 通过在 判断接收到的所述客户语音为非标准语音时， 将所述客户语音转发给座 席人员， 以提示所述座席人员进行标准发音； 然后提取接收到的所述座 席人员的标准语音的关键字； 并根据所述座席人员的标准语音的关键字 进行查询获得查询结果， 以使得所述座席人员根据所述查询结果为所述 客户提供服务。 这样客户发音不标准或用语不标准的情况下也可以进行 准确的语音识别， 提升了语音识别的准确性， 同时加快了查询速率， 提 高了工作效率。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案， 下 面将对实施例或现有技术描述中所需要使用的附图作简单地介绍， 显而易见地， 下面描述中的附图仅仅是本发明的一些实施例， 对于 本领域普通技术人员来讲， 在不付出创造性劳动的前提下， 还可以 根据这些附图获得其他的附图。 图 1 为本发明实施例提供的一种基于语音识别的查询方法流程示意 图；
图 2为本发明实施例提供的另一种基于语音识别的查询方法流程示 意图；
图 3 为本发明实施例提供的一种基于语音识别的查询装置结构框 图。
具体实施方式
下面将结合本发明实施例中的附图， 对本发明实施例中的技术 方案进行清楚、 完整地描述， 显然， 所描述的实施例仅仅是本发明 一部分实施例， 而不是全部的实施例。 基于本发明中的实施例， 本 领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他 实施例， 都属于本发明保护的范围。 实施例 1 : 本发明实施例提供了一种基于语音识别的查询方法， 如图 1所示所 述方法包括以下步骤：
101、 判断接收到的客户语音是否为标准语音。
客户拨打接入码开始访问客服系统， 通过客户终端向客服系统发送 客户语音， 客户在所述客户语音中说明客户所要进行的业务。 所述客服 系统接收到所述客户语音后会应用语音识别技术判断所述客户语音是否 为标准语音。 需要说明的是， 此处所述的标准语音既包括发音标准即客 户讲的是标准普通话； 又包括用语要标准即客户将的话中要有业务用语。
若客服系统识别出所述客户语音中有标准业务用语， 则判断接收到 的客户语音为标准语音； 否则， 判断接收到的客户语音为非标准语音。
102、 若所述客户语音为非标准语音， 则将所述客户语音转发给座 席人员， 以提示所述座席人员进行标准发音。 在判断出所述客户语音为非标准语音的情况下， 客服系统会将所述 客户语音转发给座席人员， 所述座席人员听到所述客户发音后会根据所 述非标准语音进行标准发音。
103、 接收所述座席人员的标准语音并提取所述座席人员的标准语 音的关键字。
客服系统可以接收所述座席人员的标准语音， 然后通过语音识另 'J技 术提取所述座席人员的标准语音中的关键字。
104、 根据所述座席人员的标准语音的关键字进行查询获得查询结 果， 以使得所述座席人员根据所述查询结果为所述客户提供服务。
在提取所述座席人员的标准语音中的关键字后， 所述客服系统会根 据该关键字自动查询所有业务， 获得与所述关键字相关的查询结果。 所 述查询结果包括与所述关键字相关的所有业务链接， 或与所述关键字相 关的业务页面。
若所述查询结果为与所述关键字相关的所有业务链接， 则所述座席 人员可以在所述业务链接中选择客户需要的业务链接， 点击直接进入该 业务页面为所述客户提供服务； 若所述查询结果为与所述关键字相关的 业务页面， 则所述座席人员就可以直接根据所述业务页面为所述客户提 供服务。 这样与现有技术中的全人工服务相比提高了查询速率， 进而提 高了座席人员的工作效率； 与所述语音识别服务相比， 则通过客户语音 识别与作息语音识别的结合， 提升了语音识别的准确性， 进而也提高了 查询速率。 上述步骤 101〜104所述的方法为在所述客户语音不标准的情况下进 行的查询方法； 进一步的， 如图 2所示， 所述方法还包括以下步骤：
101、 判断接收到的客户语音是否为标准语音。
若客服系统识别出所述客户语音中有标准业务用语， 则判断接收到 的客户语音为标准语音； 否则， 判断接收到的客户语音为非标准语音。
201、若所述客户语音为标准语音，则提取所述客户语音中的关键字。 若客服系统识别出所述客户语音中有标准业务用语则判断出所述客 户语音为标准语音， 则客服系统直接通过语音识别技术提取所述客户语 音中的关键字。 202、 根据所述客户语音中的关键字进行查询获得所述查询结果， 以 使得所述座席人员根据所述查询结果为所述客户提供服务。
在提取所述客户语音中的关键字后， 所述客服系统会根据该关键字 自动查询所有业务， 获得与所述关键字相关的查询结果。 所述查询结果 包括与所述关键字相关的所有业务链接， 或与所述关键字相关的业务页 面。
若所述查询结果为与所述关键字相关的所有业务链接， 则所述座席 人员可以在所述业务链接中选择客户需要的业务链接， 点击直接进入该 业务页面为所述客户提供服务； 若所述查询结果为与所述关键字相关的 业务页面， 则所述座席人员就可以直接根据所述业务页面为所述客户提 供服务。 这样与现有技术中的全人工服务相比提高了查询速率， 进而提 高了座席人员的工作效率。
上述方法可以应用在下述事例中， 全球通用户老李打 10086联系到 中国移动客服座席小张进行求助， 分为以下两情况：
A、 老李的语音为标准语音。 老李拨通电话后说 "我要查询账户信 息" ， 此时客服系统可以识别出关键字 "账户" ， 通过系统自动查询可 以获得与 "账户" 相关的所有的业务链接组成的菜单， 此时小张可以询 问老李具体需要进行的业务并选择所述菜单中对应的业务链接， 直接进 入业务页面， 这样由系统自动查询， 查询速率 4艮高。
B、 老李的语音为非标准语音， 这里所说的非标准语音有两种情况， 一种是老李说的是方言， 发音不标准； 另一种是老李说的是普通话， 但 是系统识别不出专业的业务术语。 当老李拨通用方言说 "我要查账" 或 用普通话说 "我要查账" ， 客服系统在这两种发音中都不能识别出标准 业务用语故判断老李的语音为非标准语音。 此时客服系统会将 "我要查 账"发送给小张， 小张则进行标准发音 "好的， 我来帮您查询账户概况"； 此时系统可以根据小张的发音识别出关键字 "账户" ， 通过系统自动查 询可以获得与 "账户" 相关的所有的业务链接组成的菜单， 此时小张可 以询问老李具体需要进行的业务并选择所述菜单中对应的业务链接， 直 接进入业务页面， 这样由系统自动查询， 查询速率 4艮高。 并且在老李的 语音不标准的情况下也可以应用语音识别进行语音识别查询。 本发明实施例还提供了一种基于语音识别的查询装置，如图 3所示， 所述装置包括： 判断单元 301 , 发送单元 302 , 接收单元 303 , 关键字提 取单元 304 , 查询单元 305。
判断单元 301 , 用于判断接收到的客户语音是否为标准语音。
客户拨打接入码开始访问客服系统， 通过客户终端向客服系统发送 客户语音， 客户在所述客户语音中说明客户所要进行的业务。 所述客服 系统接收到所述客户语音后会应用语音识别技术判断所述客户语音是否 为标准语音。 需要说明的是， 此处所述的标准语音既包括发音标准即客 户讲的是标准普通话； 又包括用语要标准即客户将的话中要有业务用语。
若所述判断单元 301识别出所述客户语音中有标准业务用语， 则判 断接收到的客户语音为标准语音； 否则， 判断接收到的客户语音为非标 准语音。
发送单元 302 , 用于在判断单元 301 判断所述客户语音为非标准语 音时， 将所述客户语音转发给座席人员， 以提示所述座席人员进行标准 发音。
在判断单元 301判断出所述客户语音为非标准语音的情况下， 发送 单元 302会将所述客户语音转发给座席人员， 所述座席人员听到所述客 户发音后会根据所述非标准语音进行标准发音。
接收单元 303 , 用于接收所述座席人员的标准语音。
关键字提取单元 304 , 用于提取所述接收单元接收到的所述座席人 员的标准语音的关键字。
接收单元 303可以接收所述座席人员的标准语音， 然后所述关键字 提取单元 304通过语音识别技术提取所述座席人员的标准语音中的关键 字。
查询单元 305 , 用于根据关键字提取单元提取到的所述座席人员的 标准语音的关键字进行查询获得查询结果， 以使得所述座席人员根据所 述查询结果为所述客户提供服务。
在所述关键字提取单元 304提取所述座席人员的标准语音中的关键 字后， 所述查询单元 305会根据该关键字自动查询所有业务， 获得与所 述关键字相关的查询结果。 所述查询结果包括与所述关键字相关的所有 业务链接， 或与所述关键字相关的业务页面。
若所述查询结果为与所述关键字相关的所有业务链接， 则所述座席 人员可以在所述业务链接中选择客户需要的业务链接， 点击直接进入该 业务页面为所述客户提供服务； 若所述查询结果为与所述关键字相关的 业务页面， 则所述座席人员就可以直接根据所述业务页面为所述客户提 供服务。 这样与现有技术中的全人工服务相比提高了查询速率， 进而提 高了座席人员的工作效率； 与所述语音识别服务相比， 则通过客户语音 识别与作息语音识别的结合， 提升了语音识别的准确性， 进而也提高了 查询速率。
进一步的， 所述关键字提取单元 304 , 还用于在所述客户语音为标 准语音时， 提取所述客户语音中的关键字。
若所述判断单元 301 判断出所述客户语音为标准语音， 则客服系统 直接通过语音识别技术提取所述客户语音中的关键字。
所述查询单元 305 , 还用于根据所述客户语音中的关键字进行查询 获得所述查询结果， 以使得所述座席人员根据所述查询结果为所述客户 提供服务。
在关键字提取单元 304提取所述客户语音中的关键字后， 所述查询 单元 305会根据该关键字自动查询所有业务， 获得与所述关键字相关的 查询结果。 所述查询结果包括与所述关键字相关的所有业务链接， 或与 所述关键字相关的业务页面。
若所述查询结果为与所述关键字相关的所有业务链接， 则所述座席 人员可以在所述业务链接中选择客户需要的业务链接， 点击直接进入该 业务页面为所述客户提供服务； 若所述查询结果为与所述关键字相关的 业务页面， 则所述座席人员就可以直接根据所述业务页面为所述客户提 供服务。 这样与现有技术中的全人工服务相比提高了查询速率， 进而提 高了座席人员的工作效率。 本发明实施例提供的一种基于语音识别的查询方法及装置， 通过在 判断接收到的所述客户语音为非标准语音时， 将所述客户语音转发给座 席人员， 以提示所述座席人员进行标准语音； 然后提取接收到的所述座 席人员的标准语音的关键字； 并根据所述座席人员的标准语音的关键字 进行查询获得查询结果， 以使得所述座席人员根据所述查询结果为所述 客户提供服务。 这样客户发音不标准或用语不标准的情况下也可以进行 准确的语音识别， 提升了语音识别的准确性， 同时加快了查询速率， 提 高了工作效率。 本领域普通技术人员可以理解： 实现上述方法实施例的全部或部分 步骤可以通过程序指令相关的硬件来完成， 前述的程序可以存储于一计 算机可读取存储介质中， 该程序在执行时， 执行包括上述方法实施例的 步骤； 而前述的存储介质包括： ROM、 RAM, 磁碟或者光盘等各种可 以存储程序代码的介质。
以上所述， 仅为本发明的具体实施方式， 但本发明的保护范围 并不局限于此， 任何熟悉本技术领域的技术人员在本发明揭露的技 术范围内， 可轻易想到变化或替换， 都应涵盖在本发明的保护范围 之内。 因此， 本发明的保护范围应所述以权利要求的保护范围为准。

Claims

权 利 要 求 书
1、 一种基于语音识别的查询方法， 其特征在于， 包括：
判断接收到的客户语音是否为标准语音；
若所述客户语音为非标准语音， 则将所述客户语音转发给座席人员， 以提示所述座席人员进行标准发音；
接收所述座席人员的标准语音并提取所述座席人员的标准语音的关 键字；
根据所述座席人员的标准语音的关键字进行查询获得查询结果， 以 使得所述座席人员根据所述查询结果为所述客户提供服务。
2、 根据权利要求 1所述的方法， 其特征在于， 包括：
若所述客户语音为标准语音， 则提取所述客户语音中的关键字； 根据所述客户语音中的关键字进行查询获得所述查询结果， 以使得 所述座席人员根据所述查询结果为所述客户提供服务。
3、 根据权利要求 1或 2所述的方法， 其特征在于， 所述查询结果包 括与所述关键字相关的所有业务链接， 或与所述关键字相关的业务页面。
4、 一种基于语音识别的查询装置， 其特征在于， 包括：
判断单元， 用于判断接收到的客户语音是否为标准语音；
发送单元， 用于在判断单元判断所述客户语音为非标准语音时， 将 所述客户语音转发给座席人员， 以提示所述座席人员进行标准语音；
接收单元， 用于接收所述座席人员的标准发音；
关键字提取单元， 用于提取所述接收单元接收到的所述座席人员的 标准语音的关键字；
查询单元， 用于根据关键字提取单元提取到的所述座席人员的标准 语音的关键字进行查询获得查询结果， 以使得所述座席人员根据所述查 询结果为所述客户提供服务。
5、 根据权利要求 4所述的装置， 其特征在于，
所述关键字提取单元， 还用于在所述客户语音为标准语音时， 提取 所述客户语音中的关键字；
所述查询单元， 还用于根据所述客户语音中的关键字进行查询获得 所述查询结果， 以使得所述座席人员根据所述查询结果为所述客户提供 服务。
6、 根据权利要求 4或 5所述的装置， 其特征在于， 所述查询结果包 括与所述关键字相关的所有业务链接， 或与所述关键字相关的业务页面。
PCT/CN2012/082965 2012-05-23 2012-10-15 一种基于语音识别的查询方法及装置 WO2013174092A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US13/899,110 US20130315385A1 (en) 2012-05-23 2013-05-21 Speech recognition based query method and apparatus

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN2012101617387A CN102722539A (zh) 2012-05-23 2012-05-23 一种基于语音识别的查询方法及装置
CN201210161738.7 2012-05-23

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US13/899,110 Continuation US20130315385A1 (en) 2012-05-23 2013-05-21 Speech recognition based query method and apparatus

Publications (1)

Publication Number Publication Date
WO2013174092A1 true WO2013174092A1 (zh) 2013-11-28

Family

ID=46948300

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2012/082965 WO2013174092A1 (zh) 2012-05-23 2012-10-15 一种基于语音识别的查询方法及装置

Country Status (2)

Country Link
CN (1) CN102722539A (zh)
WO (1) WO2013174092A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9471212B2 (en) 2014-03-10 2016-10-18 Htc Corporation Reminder generating method and a mobile electronic device using the same

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102722539A (zh) * 2012-05-23 2012-10-10 华为技术有限公司 一种基于语音识别的查询方法及装置
CN103795877A (zh) * 2012-10-29 2014-05-14 殷程 智能语音
CN103810211A (zh) * 2012-11-15 2014-05-21 殷程 智能语音知识库
EP3093842B1 (en) * 2014-01-06 2023-06-07 NTT DoCoMo, Inc. Terminal device, program, and server device for providing information according to user data input
CN105550235A (zh) * 2015-12-07 2016-05-04 小米科技有限责任公司 信息获取方法及装置
CN107172311B (zh) * 2017-06-22 2020-06-19 深圳市佰仟金融服务有限公司 业务评估方法及终端设备
CN109981910B (zh) * 2019-02-22 2021-08-10 中国联合网络通信集团有限公司 业务推荐方法及设备
CN110675856A (zh) * 2019-10-14 2020-01-10 苏州思必驰信息科技有限公司 用于呼叫中心的人机对话方法及装置
CN110827807B (zh) * 2019-11-29 2022-03-25 恒信东方文化股份有限公司 一种语音识别的方法及其系统
CN110956958A (zh) * 2019-12-04 2020-04-03 深圳追一科技有限公司 搜索方法、装置、终端设备及存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101656803A (zh) * 2008-08-20 2010-02-24 中兴通讯股份有限公司 一种可识别语音的座席系统及其识别语音的方法
US7672845B2 (en) * 2004-06-22 2010-03-02 International Business Machines Corporation Method and system for keyword detection using voice-recognition
CN101834900A (zh) * 2010-04-30 2010-09-15 天津大学 以地理信息为驱动的高校信息服务系统及控制模式
CN102300007A (zh) * 2010-06-23 2011-12-28 上海博路信息技术有限公司 一种基于语音识别的呼叫中心的扁平化菜单系统
CN102722539A (zh) * 2012-05-23 2012-10-10 华为技术有限公司 一种基于语音识别的查询方法及装置

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101000767A (zh) * 2006-01-09 2007-07-18 杭州世导科技有限公司 语音识别设备及其方法
JP5042799B2 (ja) * 2007-04-16 2012-10-03 ソニー株式会社 音声チャットシステム、情報処理装置およびプログラム
CN101599270A (zh) * 2008-06-02 2009-12-09 海尔集团公司 语音服务器及语音控制的方法
CN201403118Y (zh) * 2008-12-12 2010-02-10 康佳集团股份有限公司 具有翻译方言功能的装置和移动终端
CN102087854A (zh) * 2009-12-08 2011-06-08 康佳集团股份有限公司 手机智能提醒装置、方法及智能提醒手机

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7672845B2 (en) * 2004-06-22 2010-03-02 International Business Machines Corporation Method and system for keyword detection using voice-recognition
CN101656803A (zh) * 2008-08-20 2010-02-24 中兴通讯股份有限公司 一种可识别语音的座席系统及其识别语音的方法
CN101834900A (zh) * 2010-04-30 2010-09-15 天津大学 以地理信息为驱动的高校信息服务系统及控制模式
CN102300007A (zh) * 2010-06-23 2011-12-28 上海博路信息技术有限公司 一种基于语音识别的呼叫中心的扁平化菜单系统
CN102722539A (zh) * 2012-05-23 2012-10-10 华为技术有限公司 一种基于语音识别的查询方法及装置

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9471212B2 (en) 2014-03-10 2016-10-18 Htc Corporation Reminder generating method and a mobile electronic device using the same
TWI611336B (zh) * 2014-03-10 2018-01-11 宏達國際電子股份有限公司 提示產生方法、行動電子裝置及電腦可讀取媒體

Also Published As

Publication number Publication date
CN102722539A (zh) 2012-10-10

Similar Documents

Publication Publication Date Title
WO2013174092A1 (zh) 一种基于语音识别的查询方法及装置
US9117453B2 (en) Method and system for processing parallel context dependent speech recognition results from a single utterance utilizing a context database
US7940915B2 (en) Apparatus and method for determining part of elicitation from spoken dialogue data
AU2016291566B2 (en) Data driven speech enabled self-help systems and methods of operating thereof
US9542074B2 (en) Method and apparatus for enhancing an interactive voice response (IVR) system
US9270811B1 (en) Visual options for audio menu
US20110276595A1 (en) Hands free contact database information entry at a communication device
US20060287868A1 (en) Dialog system
CN103903611A (zh) 一种语音信息的识别方法和设备
US20130156170A1 (en) System and method for servicing a call
WO2015096503A1 (en) Method, device and system for associating and managing payment accounts
US6947539B2 (en) Automated call routing
US20120253817A1 (en) Mobile speech attendant access
CN107624177B (zh) 用于提高用户效率和交互性能的可听呈现的选项的自动视觉显示
US12095945B2 (en) Indicating callers for incoming voice calls on a shared speech-enabled device
CN104657492A (zh) 一种基于语音识别的设置项检索方法及系统
JP2011203349A (ja) 音声認識システム及び自動検索システム
KR100803900B1 (ko) 음성 인식 ａｒｓ 서비스 방법 및 시스템
CN106371905B (zh) 应用程序操作方法、装置和服务器
JP2016225740A (ja) 通話振り分けシステム、呼制御装置およびプログラム
US7633928B2 (en) Communication data method and system for voice applications excecutable by user equipment
US20130315385A1 (en) Speech recognition based query method and apparatus
JP2005011089A (ja) 対話装置
KR100977001B1 (ko) 고객별 맞춤 안내 기능을 가지는 자동응답시스템 및 그 방법
CN101656803A (zh) 一种可识别语音的座席系统及其识别语音的方法

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: 12877364

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: 12877364

Country of ref document: EP

Kind code of ref document: A1