CN101431573A - Method and equipment for implementing automatic customer service through human-machine interaction technology - Google Patents

Method and equipment for implementing automatic customer service through human-machine interaction technology Download PDF

Info

Publication number
CN101431573A
CN101431573A CN 200710048039 CN200710048039A CN101431573A CN 101431573 A CN101431573 A CN 101431573A CN 200710048039 CN200710048039 CN 200710048039 CN 200710048039 A CN200710048039 A CN 200710048039A CN 101431573 A CN101431573 A CN 101431573A
Authority
CN
Grant status
Application
Patent type
Prior art keywords
information
semantic
natural language
query
user
Prior art date
Application number
CN 200710048039
Other languages
Chinese (zh)
Other versions
CN101431573B (en )
Inventor
朱频频
Original Assignee
上海赢思软件技术有限公司
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

Links

Abstract

The invention provides a method realizing automatic client service using human-computer interaction technique, and an equipment realizing the method. The method includes: receiving problem imported with nature language from the user, converting the nature language into information identified by the computer using human-computer interaction technique; outputting feedback information according withthe information identified for the computer. In addition, the invention also provides the equipment realizing the method.

Description

通过人机交互技术实现自动客户服务的方法和设备 Automatic customer service through human-computer interaction method and apparatus

技术领域 FIELD

本发明涉及计算机网络技术与人机交互技术,特别是涉及一种通过计算机网络技术及人机交互技术实现智能化自动客户服务的方法及实现该方法的设备。 The present invention relates to interactive computer technology and network technology, in particular implementing the method of the intelligent automatic method is directed to a customer service implemented by a computer and network technology, and human-computer interaction.

背景技术 Background technique

以自动语音和人工服务为特点的客户服务中心在国内已经开展了多年。 Automatic and manual voice services to the characteristics of the customer service center in the country have been carried out for many years. 各种规模的企业,从银行,电信到中小企业以及政府都已经开始了这种服务。 Businesses of all sizes, from banking, telecommunications and government to small and medium enterprises have already begun such a service. 客户对这种服务形式从当初的好奇逐渐转化到习以为常。 This form of customer service from the initial curiosity gradually transformed to become accustomed. 现在,对一个企业来说,拥有客户服务中心仅仅是最低要求, 对于其客户耒说,客户服务中心必须能够提供高质量的服务。 Now, for an enterprise, with customer service centers are only the minimum requirements for their customers Lei said, the customer service center must be able to provide high quality services. 在新的 In the new

形势下,企业的客户服务中心遇到了如下的挑战: Under the situation, the company's customer service center faced challenges as follows:

1、 客服代表知识的全面性。 1, a comprehensive knowledge of customer service representatives.

随着市场竟争的加剧,企业不断推出新的产品和服务。 With the intensification of market competition, companies continue to introduce new products and services. 作为联系客户的重要部门,客户服务中心必须对企业提供的所有产品和服务都了如指掌。 As an important contact the customer department, all products and services for the enterprise customer service center must provide are well aware of. 因此,企业通常在推出新产品的前期,会对客服人员进行专门的培训,随着新产品推出速度的加快,这种培训频率也逐渐提高(成本由此增大)。 Therefore, companies often pre-launch new products, customer service staff will conduct specialized training, with the accelerated rate of new product introductions, such training frequency gradually increase (thus increasing the cost). 由于这种信息越来越多,有时甚至已经超出了客服人员能够掌握的程度,客户人员不得不在客户询问的时间,临时查询相关资料或者询问其他对此了解的同事。 As more and more of this information, and sometimes even beyond the level of customer service staff to master, the customer had to ask at the customer's time, ad hoc queries relevant information or ask other colleagues understand this. 这种临时的信息检索一方面使得客户等待更长的时间,另一方面使每次通话的成本大幅上升。 This temporary information retrieval on the one hand allows customers to wait longer, on the other hand the cost of each call to increase dramatically. 在某些情况下,客户的电话被转到不同的客服代表手中,这时客户需要不断的重复自己的问题。 In some cases, the customer's call is routed to the hands of different customer service representative, then the customer need to constantly repeat their own problems.

2、 用户数的大幅增长。 2, substantial growth in the number of users.

随着企业市场的开拓,更多的客户和潜在客户通过客服中心了解相关的产品和服务信息,这时企业原有的客户中心由于硬件系统的限制,导致客户在排队队列中等待时间过长。 With the corporate market, more customers and potential customers understand the relevant product and service information through the customer service center, then the center of the enterprise existing customers due to hardware limitations of the system, resulting in customers waiting in the queue queue for too long. 企业这时面临着一个选择, 加大设备投资(申请更多线路,提高电信设备处理能力)或者流失客户。 Enterprises then faced with a choice, increased investment in equipment (apply more lines to improve the processing capability of telecommunications equipment) or the loss of customers. 而通常情况下,客服中心的投资和再投资都非常的昂贵,并且一 And under normal circumstances, call center investment and re-investment are very expensive, and a

般需要花费3〜9月时间进行实施。 Generally it takes 3-9 months for implementation.

另一方面,随着企业各种产品和市场活动的推出,用户会在短时间之内通过客户服务中心查询和了解各种问题。 On the other hand, with the introduction of a variety of products and enterprise market activity, the user query through the Customer Service Center and understanding of the issues in a short time. 这种间断性的高峰对客服中心产生了极大的压力,如果企业对客服中心通过再投资来应对这种高峰压力,会导致非高峰时段的设备和人员冗余,这种冗余对于企业禾说是另外一种成本支出。 This intermittent peak of the call center generated a great deal of pressure, if the call center business to deal with this peak pressure through reinvestment, equipment and personnel will result in redundant off-peak periods, such redundancy for businesses Wo said another costs.

3、昼夜和假日服务。 3, day and night and holiday services.

由于工作节奏的加快,相当多用户都不方便在工作时间利用企业的客户服务中心,他们通常会选择下班以后或者利用节假日的时间。 Since the accelerated pace of work, quite a few users do not facilitate the use of enterprise customer service center during business hours, they usually choose to use the time after work or holidays. 作为企业来说,维持一个长时间的客户服务时间意味着成本的增加, 而对于客服人员本身来说,在节假日工作会带来各种各样的管理问题。 As a business, to maintain the increase in a long time means that the cost of customer service, and customer service staff for it themselves, work on holidays will bring a wide range of management issues.

在当前的应用现状下,遇到如上的挑战,企业一般别无选择,只能进行系统升级以及新客服人员的招聘,由此带来一系列的额外成本。 In the current application status, challenges encountered above, companies generally have no choice but to upgrade the system as well as recruiting new customer service staff, and the resulting series of additional costs. 但是升级及招聘都无法彻底解决客服人员对产品了解不够的问题,于是有企业投入巨资,购买和实施知识管理软件等复杂系统,由此又带来一系列的巨额成本。 However, upgrading and recruitment are not completely solve customer service problems do not know enough of the product, so companies have invested heavily, buying and implementing complex systems of knowledge management software, which brings a series of huge cost.

发明内容 SUMMARY

基于目前以自动语音和人工服务为特点的客户服务中心所存在的各种不足而提出本发明,其用于提供一种通过计算机网络技术及人机交互技术实现智能化自动客户服务的方法及实现该方法的设备。 Based on various present and artificial automated voice services to the characteristics of the customer service center of the existing problems and proposed the invention, a method for providing intelligent automated customer service is achieved through the computer network technology and human-computer interaction technology and achieve the process apparatus. .

本发明所提供的方法以及相应的设备可以用于实现客服机器人系统的构建,所述客服机器人系统或者客服机器人是一种全新的概念,它是以自然语言处理和人机交互等多种人工智能技术为基础,使用IM(即时通讯工具)或WEB网页作为表现形式的自动客服系统,在该系统中,用户可以像与人工客服对话一样与客服机器人交谈,在自然的交互过程中得到特定领域问题的解决方法。 The present invention provides a method and corresponding apparatus may be used to implement the customer to build a robot system, the robot system or call the customer service robot is a new concept, which is based on natural language processing and other interactive AI technology, the use of IM (instant messaging) or wEB page as a manifestation of automatic customer service system, in this system, the user may like to talk to a customer service, like chat with the robot, get specific problem areas during the interaction of nature in solution. 更具体而言,所述的客服机器人指的是基于网络(包括电话和移动设备)通讯的逻辑意义上的"虚拟机器人",没有自己的实体,他的功能就是为用户提供智能化自动客服。 More specifically, the robot refers to customer service is based on a "virtual robot" logical sense network (including telephones and mobile devices) to communicate without their own entity, his function is to provide users with intelligent automatic customer service. 这一系统可以应用于政府、企业以及各种事业单位中相应的向外提供服务的部门,例如,其可以用于提高政府职能部门受理来自政府、企业和市民的各类申请、提高询问和投诉的效率,显著地降低政府客服成本、减少人为错误、增强用户体验。 This system can be used in government, enterprises and institutions in various departments to provide services outside the corresponding, for example, it can be used to improve government departments handled all kinds of applications from government, business and people and to raise inquiries and complaints efficiency, significantly reducing the cost of government customer service, reduce human error and enhance the user experience.

根据本发明的第一方面,提供了一种通过人机交互技术实现自动客户服务的方法,该方法包括:接收用户输入的基于自然语言的查询信息;根据所述基于自然语言的查询信息输出查询结果。 According to a first aspect of the present invention, there is provided a method for implementing automated customer service through the human-computer interaction, the method comprising: receiving a user input query information based on natural language; query information according to an output of the natural language based query result. 具体地,该步骤中包括对作为用户输入信息的查询信息进行语义分析,并与所述预存的知识库中的内容进行语义匹配,获得与该查询信息相匹配的问题 Specifically, this step includes the query information as the user information input semantic analysis, and semantic matching with the content stored in the knowledge base, the problem of obtaining query information matches

或其相关联的答案的步骤。 Answer at step associated thereof.

优选地,当所述语义匹配获得的问题具有对应的答案,将该答案作为查询结果输出。 Preferably, when the problem of having obtained a corresponding semantic matching answer, the answer to the query result as an output. 当所述语义匹配获得的问题没有对应的答案而仅具有相关联的询问信息,则将该询问信息输出;然后接收用户针对所述询问信息而输入的基于自然语言的反馈信息;再将所述反馈信息作为所述用户输入信息来重复上述步骤,直至获得所述对应的答案和所述反馈信息为用户指示结束查询的指示信息。 The semantic matching is obtained when the problem is not only the answers corresponding to information associated with the interrogation, the output of the interrogation information; receiving feedback information and user information for the query based on the input natural language; then the feedback information as the user input information to repeat the above steps until the corresponding feedback information and said answer obtaining indication information indicating the end of the user query.

根据本发明的第二方面,提供了一种通过人机交互技术提供智能化自动客户服务的设备,该设备包括:人机交互装置,用于接收用户输入的基于自然语言的查询信息;反馈装置,用于根据所述基于自然语言的查询信息输出查询结果。 According to a second aspect of the present invention, there is provided a method of providing customer service via an automatic intelligent human-computer interaction, the apparatus comprising: a man-machine interactive means for receiving a user input query information based on natural language; feedback means for outputting information of the query results based on the query in natural language. 具体地,该装置中还包括语义匹配装置,用于对作为用户输入信息的查询信息进行语义分析,并与所述预存的知识库中的内容进行语义匹配,获得与该查询信息相匹配的问题,所述知识库包含具有一定拓朴结构的多个问题及其相关联的答案;当所述语义匹配获得的问题具有对应的答案,该语义匹配装置将该答案作为查询结果输出。 In particular, the apparatus further comprising a semantic matchmaking device, the query information for the user input information as semantic analysis, and semantic matching with the content stored in the knowledge base, the problem of obtaining query information matches the knowledge base comprises a plurality of questions having a certain topology and associated answers; matching the semantic problems when having obtained a corresponding answer, the answer to the semantic matchmaking device as the query result output.

优选地,当所述语义匹配获得的问题没有对应的答案而仅具有相关联的询问信息,则由所述语义匹配装置将该询问信息输出;然后接收用户针对所述询问信息而输入的基于自然语言的反馈信息;再根据所述反馈信息进行上述的语义匹配步骤,即以该反馈信息作为用户输入信息,对其进行语义分析并与所述预存的知识库中的内容进行语义匹配,获得与该查询信息相匹配的问题,当所述语义匹配获得的问题具有对应的答案,该语义匹配装置将该答案作为查询结果输出。 Preferably, when the semantic matching problem is not obtained only with the corresponding answer interrogation information associated, by matching the semantic query the information output means; Natural based on the query and then receiving a user for input information feedback language; then the above semantic matching step according to the feedback information, i.e. feedback information to the user as input information, and semantic analysis of its semantic content matching the pre-stored knowledge base, obtained the problem of matching the information query, matching the semantic problems when having obtained a corresponding answer, the answer to the semantic matchmaking device as the query result output. 重复上述步骤直至获得所述对应答案和所述反馈信息为用户指示结束查询的指示信息。 Repeating the above steps until the answer of the corresponding indication information and the feedback information is indicative of the end user queries.

本发明所提供方法以及系统的优点在于, Advantage of the present invention provides a method and system is that,

1) 降低企业在客户服务中心的再投资,从容应对用户数的大量增长; 1) reduce re-invested enterprises in the customer service center, calmly substantial increase in the number of users;

2) 降低企业客服中心的运行成本; 2) reduce operating costs of call center;

3) 减少客户的不愉快体验-等待时间、回答的准确性; 3) reduce customer unpleasant experience - the waiting time, the accuracy of the answer;

5) 对客户进行分析并自我更新服务能力; 5) analysis of the customer service and self-renewal capacity;

6) 低成本提供昼夜和假日服务 6) provides low-cost day and night and holiday service

采用本发明的技术方案,能够为传统客户服务中心提供问题分流支持,使传统服务中心能够专注于复杂的和高价值的客户。 The technical solution of the present invention can provide a problem for the diversion of traditional customer service center support, traditional service center to focus on complex and high-value customers. 从而在两种渠道内,企业能为用户提供了更为迅速、准确而标准的服务。 So in the two channels, companies can provide a more rapid, accurate and standard of services for users. 在低成本的条件下,为企业的高速发展提供了有效的支持。 Under conditions of low-cost, high-speed development of enterprises to provide effective support. 对于企业或者政府来说,这种领先技术的使用也为其增加了新的形象(科技和创新的形象)。 For business or government, the use of this leading-edge technology but also for its added a new image (image technology and innovation).

附图说明 BRIEF DESCRIPTION

通过阅读以下参照附图对非限制性实施例所作的详细描述,本发明的其它特征、目的和优点将会变得更明显。 A reading of the following detailed description with reference to the accompanying drawings of non-limiting embodiments, and other features, objects and advantages of the invention will become more apparent. 附图中,相同或者相似的附图标识代表相同或者相应的部件。 In the drawings, like or same reference numerals are representative of similar or corresponding parts.

图1为根据本发明的智能化客户服务设备-用户终端网络系统图; Figure 1 is a client apparatus according to the present invention, the intelligent service - user terminal network system of FIG;

图2为根据本发明的智能化客户服务设备实现智能化自动客户服务的系统方法流程图; FIG 2 is a method for intelligent automated customer service system is implemented according to a flowchart intelligent customer service apparatus according to the present invention;

图3为根据本发明的智能化客户服务设备中的知识库采用的并列式数据表示意图; 3 is employed in accordance with the customer service intelligent device of the present invention represented by the knowledge base tile data is intended;

图4为根据本发明的智能化客户服务设备中的知识库采用的树状结构数据表示意图; FIG 4 is a tree structure according to the intention of the customer service data of intelligent devices according to the present invention is employed in the knowledge base;

图5为根据本发明的一个具体实施方式通过人机交互技术提供智能化自动客户服务设备的装置框图; 5 a block diagram of the intelligent device to provide automated customer service device via a human-computer interaction according to the present invention. DETAILED DESCRIPTION;

图6为根据本发明的一个具体实施方式简化方案的变化例的智能化自动客户服务设备的装置框图; FIG 6 is a block diagram of the intelligent device automated customer service simplified variant embodiment of device according to one embodiment of the present invention;

图7为根据本发明的一个具体实施方式的智能化客户服务设备的服务器系统具体架构示意图; 7 is a client server system intelligent service device according to the present invention is a specific embodiment of particular architecture diagram;

图8为根据本发明的一个优选实施方式的智能化客户服务设备的服务器系统具体架构示意图。 FIG 8 is a schematic diagram of a device intelligent customer service preferred embodiment of the present invention, the specific architecture of the server system.

具体实施方式 detailed description

本发明可用于构建一种智能化客户服务设备(如,图l所示的智能化客户服务设备l),即本发明所述的客服机器人系统或者客服机器人平台,以实现本发明第一方面所提供的实现智能化自动客户服务的方法。 The present invention may be used to construct an intelligent customer service device (e.g., customer service intelligent device shown in FIG. L l), i.e., customer, or customer service robot system robot platform according to the present invention, to achieve a first aspect of the present invention methods intelligent automated customer service provided.

基于图1所示的智能查询系统-用户终端网络系统图,以下将详细介绍本发明第一方面所提供方法的具体过程,图2为根据本发明的智能化客户服务设备实现智能化自动客户服务的系统方法流程图。 Intelligent Query System based on FIG. 1 - FIG user terminal network system, will be described in detail hereinafter during a first aspect the present invention provides a method, Figure 2 is intelligent intelligent automated customer service customer service apparatus according to the present invention. system flowchart of a method.

本发明图1所示的智能化客户服务设备1具有一网络接口,具体地,所述网络可以是各种可以在用户终端2与智能化客户服务设备1 之间实现数据传输的网络,包括互联网,手机互联网,也可以是局域网或其他类型的网络。 Customer service intelligent device of the present invention shown in FIG 1 has a network interface, in particular, the network may be a variety of data transmission between user terminal 2 and the intelligent network customer service device 1, including the Internet mobile Internet, can also be a local area network or other type of network. 本实施方式中,所述网络接口可以用于和互联 In the present embodiment, the network interface may be used to interconnect and

9网交换数据。 9 network to exchange data.

步骤S210,用户可以向用户终端2进行基于自然语言的信息输入, 并通过互联网将该信息输入的数据传送至所述智能化客户服务设备1。 Step S210, the user terminal 2 to the user can enter a natural language based information, and transmits the data to the intelligent information input apparatus 1 via the Internet customer service.

然后进行步骤S220,智能化客户服务设备l将接收到的基于自然语言的信息输入进行语义分析,例如,提取所述自然语言中的语义关键词,将其排练组合成适于智能化客户服务设备1识别或分类检索的语句片段,然后智能化客户服务设备l将根据已建立的知识库,对经过语义分析的自然语言数据进行语义匹配,以获得所述知识库中的和该自然语言的语义相匹配的答案或者和该自然语言的语义相匹配的询问信息。 Then step S220, the intelligent device l customer service received semantic analysis of natural language based on input information, for example, extracting the semantic natural language keywords, to compositions adapted rehearsal Intelligent Service Equipment identification or classification retrieval sentence fragments, and customer service intelligent device l according to established knowledge, the semantics of data through a natural language semantic matching semantic analysis to obtain the knowledge base and the natural language the answer and the semantic or natural language matches that matches the query information.

然后执行判断步骤S230,如果在上述步骤之后所述智能化客户服务设备l已在所述知识库中获得和用户输入的自然语言的语义相匹配的答案,该智能化客户服务设备l直接将该答案作为反馈信息,执行步骤S260,将该信息通过网络输出至用户终端2,用户即可接收到反馈信息。 Then performs determination step S230, the above steps if, after said intelligent device l The obtained customer service semantics and natural language user input matches the answer the knowledge base, the intelligent service client directly to the device l the answer as the feedback information, perform step S260, the output of the information through the network to the user terminal 2, the user can receive the feedback information.

如果在上述步骤之后智能化客户服务设备l在知识库中不能获得和用户输入的自然语言的语义相匹配的答案,则该智能化客户服务设备1继续执行判断步骤S240,判断在所述知识库中是否存在和用户输入的自然语言的语义相匹配的询问信息。 If the intelligent device l can not obtain customer service semantics and natural language user input matches the answer in a knowledge base, after the above steps, the intelligent device 1 proceed to the customer service determination step S240, the determination in the Knowledge Base whether there is a semantic information and the natural language query input by the user matches. 如果存在和用户输入的自然语言的语义相匹配的询问信息,则进行步骤S241 。 If the query semantic information and the natural language input by the user matches exist, proceed to Step S241.

步骤S241,由智能化客户服务设备1从所述知识库中提取出该询问信息,并将其作为反馈信息通过互联网发送至用户终端2。 Step S241, the service device 1 extracted by the customer from the intelligent knowledge base that the interrogation message and sends it via the Internet to the user terminal 2 as feedback information.

用户可以根据用户终端2显示的所述询问信息进一步以自然语言输入信息,该输入的自然语言数据将由用户终端2通过互联网发送给所述智能化客户服务设备l,然后回到步骤S210,重复进行该智能化客户服务设备1对自然语言进行语义分析并进行语义匹配等步骤。 The user can display the user terminal 2 to the inquiry information further input information in natural language, the natural language of the input data by the user terminal 2 transmits to the intelligent device l customer service through the Internet, and then returned to step S210, repeats the intelligent customer service apparatus 1 natural language semantic analysis and semantic matching and other steps. .

如果在以上步骤之后智能化客户服务设备l在其知识库中不能获得和用户输入的自然语言的语义相匹配的答案,也不能获得和用户输入的自然语言的语义相匹配的询问信息,该智能化客户服务设备1将 If intelligent devices l can not get customer service after the above steps in its knowledge base and semantics of natural language user input matches the answer, can not get to ask the user to enter information and semantic natural language matches the intelligent customer service device 1

继续执行预定的缺省步骤S250,例如,可以根据预先设定的缺省信息输出反馈信息;或者,将所接收的无法获得语义匹配结果的自然语言输入内容通过电子邮件发送至预先设定的邮件地址,该预先设定的邮件地址例如是由专门的人工客服人员进行定时管理;还可以将人机交互界面链接至预先设定的地址,具体方式包括,通过即时通信工具MSN向客户显示一链接地址,引导客户链接到预先设定地址,该预先设定地址可以是由人工客服人员负责的客服电话,或者也可以是表现形式为MSN, QQ等即时通信工具的人工客服,由人工客服人员补充性地回答所述智能化客户服务设备l不能获得与所接收的自然语言的语义相匹配的答案的部分问题。 Continue predetermined default step S250, the example, the output of the feedback information based on the default preset information; content or natural language input, the received results can not be obtained semantic matching send a message to a predetermined e-mail address, the mail address set in advance, for example, by a live person dedicated timing manager; may also be linked to the man-machine interface address set in advance, the specific embodiment comprises, MSN instant Messenger by a link to the client display address to guide customers to link to a pre-set address, the pre-set telephone customer service address can be undertaken by a live person, or may be a form of artificial customer service MSN, QQ and other instant messaging tools, supplemented by a live person answer of the customer service intelligent device l can not be obtained with some of the problems of the received natural language semantics matches the answer.

通常情况下,当所述智能化客户服务设备l从用户终端2收到的所述对应的答案和所述反馈信息为用户指示结束查询的指示信息时,所述的流程结束。 Normally, when the smart client service from the user terminal device l 2 corresponding to the received answers and the end instruction information feedback information indicative of the user query, the process ends.

在上述过程中,可以通过自动方式或者后台管理的方式完善和增加所述智能化客户服务设备1的知识库中所存储的数据。 In the above process, can improve and increase the intelligent data repository of customer service device 1 stored in the automatic manner or by way of background management.

在智能化客户服务设备l接收用户输入的基于自然语言输入的查询信息时,这些查询信息通常因人而异,具体而言,同一种语义的查询信息,用户可以使用多种表达方式。 L service equipment receives a query-based natural language input information, which usually varies query information entered by the user in customer intelligence, in particular, with a semantic query information, the user can use a variety of expressions. 该智能化客户服务设备l中的知识库不一定能涵盖同一种语义下的多种基于自然语言的查询信息 The intelligent device l Customer Service Knowledge may not be able to cover the query information based on natural language with more under a semantic

的表达方式,而经过智能化客户服务设备l的语义分析步骤,这些表 Expression, and customer intelligence through semantic analysis step service equipment l of these tables

达方式各异的查询信息通常可以和其知识库中已存在的相应语义的 Different embodiment of the corresponding query semantic information and knowledge base may generally already present in

查询信息对应起来,从而可以进一步在所述知识库中获得和该查询信 Associating query information, and can be further obtained the query in the knowledge base channel

息相关联的答案或者相关联的询问信息。 Message asking for information or answers associated associated.

本发明的智能化客户服务设备l可以自动存储用户以各种知识库 L Intelligent Service device may automatically store the user knowledge of the present invention in various

中未有的表达方式所输入的基于自然语言的询问信息,并在经过语义分析后将其与知识库中已存储的同样语义的询问信息相关联地储存 The inputted expression No information based on natural language query, the query and the same semantic information in association with elapsed after semantic analysis stored in the knowledge base storage

起来,这种方式为追加方式的存储,在增加新的表述方式的询问信息时,并不影响知识库已经储存的任何内容,其可以丰富所述知识库涵盖的各种询问信息的表达方式。 Up in this way is stored in append mode, ask for information increase when new way of expression and does not affect any content that has been stored in the knowledge base, which can enrich the expression of various inquiries of information covered by the knowledge base. 另一方面,所存储的内容还可以用于检查所述智能化客户服务设备1对用户输入的基于自然语言的询问信息进行语义分析时是否正确,例如,可以通过人工检查上述存储的内 On the other hand, the stored content can also be used for correct semantic analysis of natural language based query information Customer service checking the intelligent apparatus 1 to user input, for example, by the manual inspection of the memory

入的基于自然语言的询问信息所关联的并不是其知识库中语义相同的询问信息,则可将该智能化客户服务设备l自动存储的询问信息删除,或者人工将其关联到另一条语义相合适的知识库中已有的询问信息上进行存储。 Based on the information not ask natural language associated with its knowledge base in the same semantics ask for information, you can ask the customer service information intelligent equipment l automatic storage of deleted or manually link it to another semantic phase appropriate knowledge base already stored on the inquiry information. 如此则可以避免下次对相同的询问信息出现不正确的语义分析。 So you can avoid incorrect semantic analysis next occurrence of the same inquiry information.

在智能化客户服务设备1与用户终端2进行信息交流的过程中,特别是其中有人工客服人员参与回答该智能化客户服务设备l不能获得与所接收的自然语言的语义相匹配的答案的问题时,可以通过该智能化客户服务设备l中相应的后台管理接口控制所述智能化客户服务设备l的知识库存储相关内容,对知识库进行更新,例如其可以存储所接收的自然语言输入内容,以及该自然语言输入内容的语义匹配结 In the process of intelligent customer service terminal device 1 2 to exchange information with the user, in particular where there is a live person involved in answering the intelligent device l can not get customer service and received a natural language semantics to match answers to questions when, through the intelligent service client device corresponding l admin interface control device of the customer service intelligent knowledge base storage content of l, update the knowledge base, which may be stored, for example, the natural language input received content , and the junction semantic matching input natural language content

果,以及其出现次数;以及所接收的无法获得语义匹配的答案和语义匹配的询问信息的自然语言输入内容及其由人工客服人员相应回复的反馈信息;另外,还包括存储一些涉及专业领域的语言词汇。 Fruit, their occurrence; and the received feedback information can not be obtained, and semantic matching semantic matching answers natural language query information input by the content and the corresponding reply to a live person; additionally, further comprising storing a number of fields involving professional language vocabulary. 通常情况下,知识库所存储的一项或多项内容任选地可以通过替换或者追加方式更新所述知识库。 In general, knowledge of one or more of the stored content of the knowledge base can be optionally updated by replacing or adding manner.

具体地,所述智能化客户服务设备1中的知识库可以采用存储方式为并列式的数据表,这种情况下,其分为两个数据表,如图3所示, 其中,第一数据表用于存储和基于自然语言的特定问题语义相匹配的预设答案,例如,预设答案al和问题ql相关联,预设答案a2和问题q2 相关联等等,这些和用户基于自然语言输入的问题相匹配的答案在数据库中的关系是并列的;第二数据表用于存储和特定自然语言的语义相匹配的预设询问信息,例如,预设询问信息al,和问题ql,相关联, 预设询问信息a2,和问题q2,相关联等等,这些和用户基于自然语言输入的问题相关联的询问信息在数据库中的关系也是并列的。 In particular, the intelligent service client apparatus 1 can be used in the knowledge base storage into parallel data table, in this case, which is divided into two data tables, shown in Figure 3, wherein the first data table for storing natural language and semantics of the particular problems that match preset answers, e.g., default questions and answers al associated ql, q2 preset answers a2 and associated problems, etc., based on the natural language input and user match answers the question in the database relations are juxtaposed; preset query semantic information of the second data table for storing natural language and a particular match, for example, Al preset information inquiry, and QL problems, associated the preset asking for information a2, and issues q2, etc. associated with these issues and ask for information based on user input natural language associated relationships in the database are also parallel. 此外,所述知识库也可以采用树状结构形式的数据表,这种情况下,该数据表中对特定自然语言根据其语义预先分类并逐级排列,如图4所示,该树状数据表是将用户基于自然语言输入的问题逐级地进行分类排列,例如,用户基于自然语言输入的问题经过语义分析并经 Furthermore, the knowledge base may be employed in the form of a tree structure of the data table, in this case, the data table previously classified according to its specific natural language semantics stepwise arrangement shown in Figure 4, the tree data the table is arranged to classify the user step by step based on the natural language input, for example, based on a user input through the natural language semantic analysis and dried

过语义匹配为Q1类问题时,在该类问题将存在相应的若干条询问信息,可继续用询问或提示的方式使用户进一步回答,例如分为Qll、 Q12、 Q13等问题对用户进行提问,从而在最后得到和用户基于自然语言输入的问题语义相关联的答案。 Q1 is excessively based semantic matching problems, such problems in the presence of the respective interrogation several pieces of information, can continue to prompt manner enables the user to ask or answer further, for example into Qll, Q12, Q13 and so on for the user to ask questions, in order to finally get answers to natural language input and user-based semantics associated problems. 当然,这种树状结构的数据表也可以在用户输入的基于自然语言的问题具有直接相关联的答案时直接得到答案。 Of course the answer directly, data tables Such a structure can also answer questions directly associated natural language-based user input.

以下将通过图5所示的通过人机交互技术提供智能化自动客户服务设备l的装置框图,并结合一个实施例对本发明第二方面提供的实现自动客户服务的设备进行具体的描述。 Will be provided as shown in FIG. 5 by a block diagram of the intelligent human-computer interaction device l automated customer service device and device in connection with one embodiment to achieve automated customer service will be specifically described embodiments of the second aspect of the present invention provides.

如图5所示,根据本发明的一个具体实施方式的通过人机交互技术提供智能化自动客户服务的智能化客户服务设备l主要适用于提供咨询、问讯或者类似业务的单位或者单位中负责相应职能的机构。 As shown, automated customer service 5 intelligent computer interaction technique in accordance with a particular embodiment of the present invention, customer service intelligent device primarily suitable for providing advice l, inquiries or similar business unit or units is responsible for the corresponding agency functions. 该智能化客户服务设备1主要包括一人机交互装置51以及一反馈装置 Customer Services The intelligent device 1 mainly includes a man-machine interaction means 51, and a feedback means

52, 优选地,该智能化客户服务设备5中还可以包括一后台管理装置53。 52, preferably, the customer service intelligent device 5 may further include a background management apparatus 53.

人机交互装置51主要用于将所接收到的来自用户终端2的数据, 例如用户在用户终端2上用上述基于windows的微软拼音输入法输入的文字,或者用户在用户终端2上通过麦克风而输入的声音等通过人机交互技术转换成供计算机识别的信息;具体包括将所述数据中所包含的自然语言转换成适于该智能化客户服务设备l识别或分类检索的信息,例如,将用户的声音输入转化为计算机可以进行语义分析的数据。 Human-machine interaction means 51 is mainly used for the terminal 2 of the data received from the user, for example, a user on the user terminal 2 based on the above-described character input windows Microsoft Pinyin input method, or the user on the user terminal 2 through the microphone by converting the input audio and other interactive computer technology to for identification information; comprises converting the natural language data contained in the client device to the intelligent l identification or classification information retrieval service is adapted to, for example, user's voice input into a computer for data semantic analysis. 另一方面,人机交互装置51还用于将来自反馈装置52的反馈信息通过互联网发送到用户终端2。 On the other hand, human-machine interaction means 51 for further feedback information from the feedback device 52 is transmitted to the user terminal 2 via the Internet.

反馈装置52通常包括一知识库存储装置521和语义匹配装置522,其中知识库存储装置521具体的表现形式为各种格式的关系型数据库例如DB2、 Oracle、或者MySQL, Microsoft SQL Server 2000 with SP3/SP4等等。 Feedback means 52 generally comprises a knowledge base storage unit 521 and the semantic matchmaking device 522, wherein the specific knowledge base storage means 521 forms a variety of formats, for example, relational databases DB2, Oracle, or MySQL, Microsoft SQL Server 2000 with SP3 / SP4 and so on. 其中存储和特定的自然语言的语义相匹配的答案,以及和特定的自然语言的语义相匹配的问题。 The semantics and storage problems which specific natural language matches the answer, as well as specific natural language and matches. 语义匹配装置522用于对自然语言进行语义分析并进行语义匹配,该语义匹配装置522将上述通过人机交互装置51转换的各种数据(即用户通过用户终端2输入并通过互联网发送至人机交互装置51的自然语言数据)进行语义分析, 例如,提取其中的关键词,然后将其与知识库存储装置521中所存储的和特定的自然语言的语义相匹配的答案或者问题进行匹配,并将匹配结果反馈至所述人机交互装置。 The semantic matchmaking device 522 for semantic analysis of natural language and semantic matching, the semantic matchmaking device 522 (i.e., the second input via the user data is converted by the above-described various man-machine interaction means 51 and sent via the Internet to a terminal man-machine Natural language data interaction device 51) semantic analysis, for example, to extract the keyword, then the knowledge base storage unit 521 and stored in a specific natural language semantics answer matches or matching problem, and the matching result is fed back to the interactive device.

具体地,输出反馈信息时,反馈装置52是由语义匹配装置522根据已建立的知识库存储装置521中的数据库,对经过人机交互装置51 转换的自然语言数据进行语义匹配,以获得和该自然语言的语义相匹配的答案并将其反馈至人机交互装置51。 Specifically, when the output feedback information, the feedback means 52, a natural language interactive means 51 through the data conversion is performed by a semantic matching semantic matchmaking device 522 according to the established knowledge base storage means 521 in the database, and to obtain the semantics of the natural language answer matched and fed back to the interactive device 51. 如果所述语义匹配装置522 没有获得和该自然语言的语义相匹配的答案,则其进一步在所述知识库存储装置521的数据库中匹配和该自然语言的语义相匹配的询问信息,如果匹配到和该自然i吾言的语义相匹配的询问信息,则输出该询问信息至人机交互装置51,并由人机交互装置51通过互联网将询问信息发送至用户终端2,用于向用户提问;然后用户可以通过用户终端2 输入用于回答所述询问信息的自然语言,并经由互联网发送至人机交互装置51,语义匹配装置522接着对经过人机交互装置51处理的反馈自然语言数据进行语义匹配;重复上述步骤,直至用户终止在用户终端2上输入自然语言。 If the semantic matchmaking device 522 is not obtained and the semantics of the natural language matches the answer, which further interrogation information and semantic matching the natural language matches in the knowledge base storage means database 521, if a match to the inquiry information and semantic nature Mark me i match, then outputs the interrogation information to the interactive means 51, by the interactive device 51 via the Internet inquiry information to the user terminal 2, to ask the user; the user may then input through the user terminal 2 for answering said natural language query information, and transmitted to the interactive device 51 via the Internet, then the semantic matchmaking device 522 through the feedback data interactive natural language processing apparatus 51 semantically matching; repeating the above steps until the user terminates the natural language input in the user terminal 2.

上述的知识库存储装置521还可存储人机交互装置51所接收的自然语言输入内容及其经过人机交互装置51处理以及语义匹配装置522匹配的结果,用于更新该知识库存储装置521中的知识库,所迷语义匹配结果包括和用户在用户终端2上输入的自然语言的语义相匹配的答案或者询问信息。 Natural language input contents of the knowledge base storage means 521 may also store the received interactive means 51 and processing means 51 through the man-machine interaction, and matching the results of the semantic matchmaking device 522, for updating the knowledge base storage means 521 knowledge base, the fans semantic matching results include semantic and user natural language input from the user terminal 2 matches the query answers or information. 由于用户在用户终端2上输入的问题多种多样,例如,其中一部分和应用该智能化自动客户服务设备1的单位中相应职能部门所负责的咨询业务没有直接关系,或者虽然和该职能部门的咨询业务有关但 Since the user in question on user input terminal 2 variety, for example, part of the application and intelligent automated customer service equipment unit 1 of the corresponding functional departments are responsible for consulting services not directly related to, or though and the functional departments consulting services related to but

是相应的问题在知识库存储装置.521的知识库中尚未有相应的或者相关的答案,这种情况下,该智能化自动客户服务设备5将不能获得与所接收的自然语言的语义相匹配的答案,甚至也无法根据所接收的自然语言的语义向用户反馈询问信息。 The corresponding knowledge base storage means .521 problem has not been a corresponding knowledge base or associated answers, in this case, the intelligent automatic customer service can not be obtained with device 5 to the received natural language semantics matches the answer, not even asking for information feedback to the user based on the received semantic natural language. 根据本发明的优选方案,当上述反馈装置52没有获得与所接收的自然语言的语义相匹配的答案和询问信息时,其可以执行预定的缺省指令,通过所述人机交互装置41 向用户终端2反馈代偿性的内容,例如,反馈装置52可以根据预先设定的缺省信息输出反馈信息;或者,将所接收的无法获得语义匹配结 According to a preferred embodiment of the present invention, when the feedback means 52 and the answer is not obtained semantic information inquiry received matches the natural language, which may be performed a predetermined default command, the human-machine interaction means 41 by the user compensatory feedback terminal 2 content, for example, the feedback device 52 may output feedback information based on the default information set in advance; or the received junction semantic matching can not be obtained

预先设定的邮件地址例如是由专门的人工客服人员进行定时管理;还可以将人机交互界面链接至预先设定的地址,具体方式包括,通过即时通信工具MSN向客户显示一链接地址,引导客户链接到预先设定地址,该预先设定地址可以是由人工客服人员负责的客服电话,或者也可以是表现形式为MSN, QQ等即时通信工具的人工客服,由人工客服人员补充性地回答所述反馈装置52不能获得与所接收的自然语言的语义相匹配的答案的部分问题。 A preset mail address, for example, by a live person dedicated timing manager; may also be linked to the man-machine interface address set in advance, including the specific embodiment, MSN displaying a link address to the client via the IM tool, the guide customer link to a pre-set address, the pre-set telephone customer service address can be undertaken by a live person, or may be a form of artificial customer service MSN, QQ and other instant messaging tools from a live person answers supplementally the feedback means 52 part of the problem can not be obtained and the received natural language semantics matches the answer.

为了使本发明的智能化自动客户服务设备l在使用过程中不断完善和增加知识库存储装置521中所存储的知识库内容,在智能化自动客户服务设备1与用户终端2进行信息交流的过程中,特别是其中有人工客服人员参与回答反馈装置52不能获得与所接收的自然语言的语义相匹配的答案的问题时,知识库存储装置521可以存储相关内容, 用于更新知识库,例如其可以存储所接收的自然语言输入内容,以及该自然语言输入内容的语义匹配结果,以及其出现次数;以及所接收的无法获得语义匹配的答案和语义匹配的询问信息的自然语言输入内容及其由人工客服人员相应回复的反馈信息;另外,还包括存储一些涉及专业领域的语言词汇。 In order to make the present invention intelligent automated customer service devices continue to improve and increase knowledge l knowledge base storage unit 521 in the content stored during use, intelligent automatic exchange of information service device 1 and the client terminal 2 of the user process in particular wherein there are involved in a live person answering questions can not be obtained feedback means 52 and the received natural language semantics matches the answer, knowledge base storage means may store the content 521 for updating the knowledge base, which e.g. natural language input may store the received content, and semantics of natural language input matches the content, and its number of occurrences; and natural language input inquiry contents of the information received can not get answers semantic matching and semantic matching and by the artificial customer service feedback information corresponding reply; additionally, further comprising relates to the field store some specialized language vocabulary. 通常情况下,知识库存储装置521所存储的一项或多项内容任选地可以通过覆盖或者追加方式更新所述知 Typically, a 521 stored in the knowledge base storage means to one or more of the content optionally may additionally known manner by covering or update

识库存储装置521中的知识库。 Knowledge base storage means 521 in the knowledge base.

为了实现上述的优选方案,本发明的智能化自动客户服务设备l 可以包括一后台管理装置53,其用于完善和增加知识库存储装置521 中所存储的知识库内容。 To achieve the above preferred embodiment, the customer service intelligent automatic apparatus according to the present invention may comprise l a background management device 53, for increasing and improving the knowledge base knowledge base storage content stored in the device 521.

所述后台管理装置53能够帮助使用该智能化自动客户服务设备1 的用户管理和部署该智能化设备的功能。 The back-end management device 53 can use the help of intelligent automated customer service device user management and deployment capabilities of the intelligent device 1. 具体地,所述后台管理装置53可以是,例如,连接到反馈装置52的电脑终端,它支持UNIX、Linux、 Windows等操作系统,管理人员可以通过基于WEB的管理后台方便地管理的该智能化自动客户服务设备1的功能,如调整该智能化自动客户服务设备l的客服系统的表现形式等操作。 In particular, the background manager 53 may be, for example, a computer terminal connected to the feedback means 52, which supports UNIX, Linux, Windows and other operating systems, WEB-based management can easily manage management background of the intelligent customer service functions of an automatic apparatus, such as adjusting the form of intelligent devices l automated customer service customer service system operation. 具体包括:1)、客服知识库的增加、删除和修改;2)、客服知识库的批量导入,例如本发明采用的操作系统可支持多种方式的数据导入(Excel、 Access、文本文件等),并且可以选择是覆盖/追加等选择项;3)、知识库的处理策略,当数据库中没有与用户的提问相关的问题时,使用何种策略应对, Including: 1), increasing customer knowledge base, deletions and modifications; 2), bulk import customer knowledge base, such as an operating system of the present invention may be employed in a variety of ways to support data import (Excel, Access, text files, etc.) and you can choose to cover / append and other options; 3), processing strategy knowledge base when not in the database associated with the user's asking questions when, what strategy to use to deal with,

可选择缺省回答、Email给人工客服、转接人工客服(MSN机器人支持) 等多种方式;4)、管理(增加和修改)机器人的欢迎语;5)、管理(增加和修改)机器人的缺省回答;6)、修改人工客服的信息;7)、 日志分析功能,可以查看系统的详细使用情况,具体功能如下:问题内容,回答内容,IP地址,提问时间,响应时间;针对某IP察看该IP 提问列表以及回答内容;通过日志分析,系统管理员能标记错误回答, 使机器人再次遇到相同问题时可以规避错误答案;提供问题被提问次数的排名;提供筛选功能,能筛选出正确回答、错误回答、供选择的回答、无法回答等情况。 Choose the default answer, Email to a human customer service, call forwarding artificial (MSN robot support) and other methods; 4), management (additions and modifications) robot greeting; 5), management (additions and modifications) robot the default answer; 6), modify the information artificial customer service; 7), log analysis function, you can view the detailed use of the system, specific functions are as follows: content, answer content, IP address, question time, response time; for a certain IP View the list of questions and answers IP content; the log analysis, system administrators can mark the wrong answer, the robot can avoid the wrong answer again encountered the same problem; the problem is to provide the number of ranking questions; provides filtering capabilities, can filter out the right answer, wrong answer, alternative answer, could not answer and so on. 同时提供提问时间篩选、IP地址筛选、问题内容筛选;8)、分析机器人无法回复的问题列表,添加正确的问题和答案到知识库中,不断的丰富知识库,从而提高机器人的智能;9)、 增加专业词汇,丰富机器人的专业领域知识;IO)灵活的管理人员权限设置。 Question time while providing filtering, IP address filtering, content filtering problem; 8), analyze the robot can not reply to a list of questions, the right questions and answers add to the knowledge base, and continue to enrich the knowledge base, thereby enhancing the intelligence of robots; 9 ), increase vocabulary, rich domain expertise of the robot; IO) flexible management of permissions.

上述后台管理装置53并不是必需的,该装置的作用是在本发明的智能化自动客户服务设备l能够针对客户输入的自然语言进行自动回复的基础上,进一步完善本发明所提供设备的智能化程度,但是在所述智能化自动客户服务设备1中的知识库已经较为完善的情况下,可 Necessary, the action of the management apparatus 53 is not the background of the device is based on the intelligent automated customer service device l according to the present invention can be automatically reply for the customer to enter a natural language, the present invention is to further improve the intelligent device provided degree, but in the intelligent automated customer service device 1 in the Knowledge Base have been more perfect case,

以不使用后台管理装置53。 Not to use the back office device 53. 换言之,只需要有所述的反馈装置52以及人机交互装置51已经可以完成本发明的基本方案,或者,使用后台管理装置53使所述知识库较为完善时,可以去掉后台管理装置53。 In other words, only to have the feedback means 52 and the interactive device 51 can be completed in the basic embodiment of the present invention, or using a background knowledge management apparatus 53 when the better, the management apparatus 53 can remove the background. 因此, 作为本发明实施例一个变化例,所述的智能化自动客户服务设备可以不包括所述的后台管理装置。 Accordingly, the present invention is a variant embodiment of the embodiment, the intelligent automatic service apparatus may not include a customer management apparatus according to a background. 如图6所示,该智能化客户服务设备5, 主要包括一人机交互装置51以及一反馈装置52,所述反馈装置52中还包括知识库存储装置521以及语义匹配装置522。 6, the customer service intelligent device 5, including a human-machine interaction means 51, and a feedback device 52, the feedback device 52 further includes a knowledge base storage means 521 and the semantic matchmaking device 522. 该简化实施例的智能化自动客户服务设备5,中,上述各装置可以完成的功能和前一个设备的实施例相同。 The same as Example 5 intelligent automated customer service device, in each device can be completed before a functional device and a simplified embodiment of the embodiment.

此外,作为本发明实施例的另一个变化例,以上实施例所述的知识库存储装置521或者语义匹配装置522均可以独立于反馈装置52,而由后台管理装置对其内容或者功能加以控制。 Further, as another variation of embodiment of the present invention, the embodiment, according to the above embodiment knowledge base storage means 521 or the semantic matchmaking device 522 may be each independently in the feedback means 52, and the control apparatus to be managed by the background function or its content. 在实际应用时,本领域技术人员可以根据需要将本发明实施例中所描述的各种装置拆分或者结合后加以应用。 After be applied in practical applications, those skilled in the art may be necessary to split the various devices described in the embodiments of the present invention or in combination.

以下将详细介绍根据上述内容利用服务器系统等具体设备构建形成本发明所述客服机器人的技术方案,在根据本发明提供的方法和设备实施上述客服机器人方案之前,通常需要进行如下的准备工作: The following describes in detail the present invention constructed in accordance with the form aspect of customer service robot, prior to the above-described embodiment embodiment customer robot apparatus and method according to the present invention provides, as typically required above preparation using specific device server system:

硬件环境的准备 Ready hardware environment

一台或一台以上个人电脑终端服务器,其用于作为数据库以及应用程序服务器;在采用一台以上服务器的情况下,数据库服务器和应用程序服务器可以分开设置,以便于提高性能以及方便网络架构。 One or more of a personal computer terminal server, which is used as a database and an application server; in the case where more than one server, database server and application server may be provided separately, in order to improve performance and facilitate network architecture.

上述的服务器可以采用各种品牌和配置的服务器,例如,可以采用Dell、 HP、 IBM、联想、浪潮等品牌的厂商所提供的各种型号的服务器。 The aforementioned server can use a variety of brands and configurations of the server, for example, various models of Dell, HP, IBM, Lenovo, wave and other brands of manufacturers offer server can be used. 软件环境的准备 Preparation software environment

与硬件环境相应,服务器上的操作系统也可以有多种选择,例如, Corresponding to the hardware environment, the operating system on the server may have many options, e.g.,

其可以采用Windows作为才喿作系统,同样也可以采用Unix、 Linux、 Linux、 FreeBSD、 Solaris等等作为操作系统。 It can be used as a Windows system as the only Qiao, also can be used Unix, Linux, Linux, FreeBSD, Solaris and so on as the operating system. 通常,用于实现本发明技术方案的服务器可以安装Microsoft Windows 2003 Enterprise操作系统,并安装其它如防火墙等安全和性能调节软件。 Typically, the server technology used to implement the present invention can be installed Microsoft Windows 2003 Enterprise operating system, and install security and other issues such as performance tuning software firewall. 数据库服务器还需要安装Microsoft SQL Server 2000 with SP3/SP4。 The database server also need to install Microsoft SQL Server 2000 with SP3 / SP4. 在客服软件实施时通常需要使用到数据库的管理密码。 Typically use password management software customer database at the time of implementation.

此外,本发明的技术方案所涉及的数据库为一个泛指的概念,其可以是各种格式的关系型数据库,而不局限于某种特定格式的数据库。 Further, the database aspect of the present invention refers to a concept which can be a relational database in various formats, but not limited to a specific database format. 具体地,在应用本发明的技术方案时,可以选择商业数据库,例如DB2、 Oracle等等,也可以选择免费的数据库,例如MySQL等等。 Specifically, when applying the technical solution of the present invention can be selected commercial databases, such as DB2, Oracle, etc., you can also choose a free database, such as MySQL and so on.

网络环境的准备 Preparing the Network Environment

确保根据本发明所提供的技术方案而设置的服务器相互之间通过内部网连接(在采用一台以上服务器时),例如,通过交换机将各服务器组成局域网,并且这些服务器都分配有固定的内网IP地址。 According to the server to ensure that the present invention is provided and connected to each other is provided (when using more than one server), e.g., by the respective switch intranet LAN servers, and these servers are assigned a fixed network IP addresses. 由于要提供互联网WEB服务,应用程序服务器必须有到外网的出口(预先保留一出口带宽)。 Because of the Internet to provide WEB service, application server must have to exit the external network (pre-reserved bandwidth of an outlet). 并且,根据本发明的技术方案而构建的客服机器人在实施时需要使用到外网地址和映射的端口号(例如80端口)。 Further, according to the aspect of the present invention constructed embodiment when the robot customer need to use external network address and port number (e.g. port 80) mapping.

本领域技术人员应当理解,上述的各种准备工作仅仅是用于示范性地对本发明的具体实施方式进行说明,而不是用于限制本发明。 Those skilled in the art will appreciate, the above-described various preparations are merely exemplary for the specific embodiment of the present invention will be described embodiments, and are not intended to limit the present invention. 在实际应用过程中,技术人员可以根据实际需要对所述硬件环境、软件环境、网络环境中的任一项或者任多项内容作适当调整或者补充。 In practical applications, the skilled person can be made according to actual needs of any one of the hardware environment, software environment, the network environment or any or several of the adjusted appropriately added. 例如,在上述的软件环境之外,本领域技术人员还可以根据需要安装支撑软件,或者,在本发明技术方案基础上对所述通过各种服务器构建的客服系统进行安装和配置,具体地,所述的安装支撑软件可以包括, For example, in addition to the above software environment, those skilled in the art may also be required to support the installation of software, or installation and configuration of the customer service system constructed by a variety of servers based on the aspect of the present invention, in particular, the software may include a mounting support,

18例如JDK1.5, Resin-Pro, JDBC组件等等,安装客服系统包括智能客服引擎、自然语言处理引擎、管理控制台等。 18 For example JDK1.5, Resin-Pro, JDBC components, etc., installation of customer service system includes an intelligent customer engines, natural language processing engine, management consoles.

此外,所述的硬件环境也不局限于上述用于示例的各种硬件环境,例如,在某单位中实施根据本发明的技术方案而构建的客服机器人时,所涉及的各种应用程序服务器、数据库服务器等均可整合在该单位原有的各种型号的服务器中,而不限定于本发明所列举的各种服务器。 In addition, the hardware environment is not limited to the above example hardware environment used, for example, customer service robot According to the embodiment of the present invention is constructed in a unit, the various application servers involved, the database server can be integrated in the unit and other various types of original server, without limiting the present invention to various servers listed. 或者,用于构建根据本发明的客服机器人的各种应用程序服务器、数据库服务器是由单独购置的服务器构成,但是这些服务器也可以与该单位原有的服务器或者服务器系统配合使用,例如,所述单位 Alternatively, constructed according to the various application servers according to the present invention, the customer service robot, the database server is made to purchase a separate server, these servers can also be used with the units of the original server or server system, for example, the unit

原有一台互联网WEB服务器,其具有连接internet的出口,该服务器用于向外界展示所述单位的网站,在这种情况下,通常由技术人员根据网络的实际情况将根据本发明的技术方案而构建的WEB客服机器人的统一资源定位符(URL )通过例如JavaScript脚本的方式集成到该单位已有的网站中去。 An original Internet WEB server having an outlet connected to the internet, the server is configured to show to the outside of the unit site, in this case, usually by the skilled person according to the actual situation of the network will According to the present invention Construction of the uniform resource locator (URL) WEB customer service robot to integrate into existing websites of the unit by way of example, JavaScript scripts.

图7示出了本发明的一个具体实施方式的客服机器人系统架构示意图。 FIG. 7 shows a specific embodiment of the present invention, a schematic diagram of the architecture of customer service robot system. 该客服机器人通常的应用环境可以是提供咨询、问讯或者类似业务的单位,在此假定其用于某单位中与公众进行沟通的一具体部门。 The customer service robot usual application environment can provide advice, information desk or similar business units, it is assumed here that a specific department to communicate for a unit with the public.

在进行所述客服机器人系统的构建时,所采用的服务器等具体设备可以与上述内容中所介绍的装置——对应,例如,在图1或者图2 中所示的人机交互装置71通常可以是一个负责与用户交换数据的服务器,例如一个WEB服务器;在图5或者图6中所示的语义匹配装置522 通常可以是一个专门的具有智能客服引擎的服务器;而知识库存储装置通常可以是一个用于存储的知识库服务器。 When constructing the robot system during the call, the specific servers and other devices may be employed with the apparatus described in the above - corresponds, for example, may generally be in human-computer interaction device 1 shown in FIG. 2 or FIG. 71 It is responsible for a user to exchange data with the server, a WEB server, for example; in the semantic matchmaking device shown in FIG. 5 or FIG. 6 522 generally may be a dedicated engine intelligent call servers; and the knowledge base storage means may typically be a repository server for storage. 当然,为合理利用服务器等硬件设备,上述图5或者图6中所示意的各个装置均可以整合在一个服务器中。 Of course, for the rational use of servers and other hardware devices, each of the above-described FIG. 5 or the apparatus illustrated in FIG. 6 can be integrated in one server. 另一方面,也同样可以由一个以上服务器构成上述内容所述的一个装置,举例而言,上述的反馈装置52可以是由一个知识库服务器和一个智能引擎服务器构成。 On the other hand, also more than one server may be composed of a device according to the above, for example, the above-described feedback means 52 may be constituted by an intelligent knowledge base server and a server engine. 因此,以下所述的各种服务器仅仅是对本发明一个具体应用方案的举例,在实际应用中将可以根据上述提示进行各种调整。 Thus, according to the following various server applications is merely one specific example of the present invention, can be variously adjusted according to the above application will prompt practical.

如图7所示,根据本发明的客服机器人平台体系架构中采用的服 7, the customer service robot platform architecture of the present invention is employed in accordance with

务器系统包括:知识库服务器71、智能客服引擎服务器72、以及互联网WEB服务器73,这些服务器构成智能化自动客户服务设备l。 Service system comprising: a knowledge base server 71, the customer service intelligent engine server 72, and Internet WEB server 73, which constitutes the intelligent automated customer service server device l. 其中, 分别由三台服务器执行知识库(数据库)存储、智能客户服务以及和互联网进行数据交换的功能。 Among them, were executed Knowledge Base (database) storage, intelligent customer service as well as by the three servers and Internet data exchange function. 并且,如图7所示,这些服务器之间通过线缆或者内部网互连,彼此之间可以进行相互之间的通信,任意两个服务器之间均可以相互交换数据。 As shown in FIG 7, interconnected by a cable or an intranet, can communicate among each other between each of these servers, data can be interchanged between any two servers.

用户可以通过用户终端2经由互联网登录到上述该政府部门的网站,即通过互联网连接到图7所示的互联网WEB服务器73,并通过WEB交互方式或者各种即时通讯工具,例如MSN, QQ等方式与上述知识库服务器71、智能客服引擎服务器72、以及互联网WEB服务器73 所组成的服务器系统进行信息交流。 Users can log in via the user terminal 2 via the Internet to the above the government's website, WEB server connected to the Internet as shown in Figure 773 through the Internet, and through the WEB interactively or various instant messaging, such as MSN, QQ, etc. 71, intelligent call server engine 72, and the Internet server WEB server system 73 composed of the exchange of information with the above-described knowledge base server.

具体地,用户以所述的WEB交互方式和该服务器系统构建的智能化自动客户服务设备l进行信息交流时,用户终端2只需要拥有和互联网WEB通信协议相应的网络浏览器,例如微软公司的IE,这种浏览器通常集成于常见的操作系统如Windows中,不需要安装或者修改任何软件,即可以方便地和所述智能化自动客户服务设备1进行信息交流。 In particular, according to the user interaction and the WEB server client system constructed intelligent automatic service apparatus when l exchange of information, the user terminal 2 need only have a corresponding communication protocol and WEB Internet web browser, such as Microsoft IEs, such browsers typically integrated in a common operating systems such as Windows, no need to install any software or modification, i.e. can be easily service device 1 and the information exchange intelligent automatic client. 该用户终端2能够支持现有的多种操作系统,例如,Windows, Unix, Linux等。 The user terminal 2 is able to support the existing variety of operating systems, for example, Windows, Unix, Linux and so on.

对于习惯使用即时通信工具的用户,根据本发明的服务器系统构建的智能化自动客户服务设备l同样也可以支持即时通信工具的使用,例如MSN, QQ等常见的即时通信工具。 For users accustomed to using instant messaging tools, server system according to the present invention is to build intelligent automated customer service equipment l can also support the use of instant messaging tools such as MSN, QQ and other popular instant messaging tools. 在这种情况下,所述客服机器人通常以使用该即时通信工具的另一端的聊天对象的形式而出现,这是一种更为轻松和随意的交流环境,能够让用户部分消除面对重要部门时的焦虑和紧张。 In this case, the customer service robot is usually in the form of chatting with the other end of the use of instant messaging tools emerged, which is a more relaxed and casual exchange environment, allowing users to eliminate part of the face of an important sector anxiety and tension when.

以上两种和用户终端2的信息交流方式可以单独使用,也可以合并使用,在实际应用时可以根据具体情况加以选择,不管何种方式均可用于完成本发明的技术方案。 Two or more user terminals to exchange information and Embodiment 2 may be used alone or in combination, may be selected in practical applications depending on the circumstances, no matter how complete the art may be used in the present invention.

在应用根据本发明的服务器系统所构建的智能化自动客户服务设备l时,用户如果需要同上述的某单位中设置有客服机器人系统的具体部门进行信息交流,例如咨询和该具体部门业务相关的信息,可 In the application according to the intelligent automatic client-server system according to the present invention is constructed L service device, if the user needs to set a specific customer service departments of a robot system with a unit of the above-described information exchange, such as counseling and the sector-specific service-related information,

以通过用户终端2以自然语言输入问题,假定用户终端2是通过即时通信工具MSN和智能化自动客户服务设备1进行信息交流,则上述以自然语言输入的信息将通过即时通信工具MSN经由互联网发送到所述服务器系统中的互联网WEB服务器73,所述智能化自动客户服务设备1通过该互联网WEB服务器73接收所述以自然语言输入的问题。 To the user terminal 2 to input natural language question, assuming that the user terminal 2 is an exchange of information through MSN instant messaging tools and intelligent automated customer service equipment, the above information will be sent via the Internet to natural language input through MSN instant messaging tool to the Internet server WEB server system 73, the intelligent automatic customer service device 173 receives the natural language question input by the Internet server WEB.

具体地,上述的客户基于自然语言输入问题的方式可以是向用户终端2通过各种输入法,例如在键盘上使用微软拼音输入法进行文字输入,也可以是向用户终端2通过麦克风而进行的声音输入,在声音输入的情况下,将涉及到对输入的声音进行识别以判断其语义的过程,因其属于成熟的现有技术,在本发明中将不再详细叙述。 In particular, the above-described manner based on customer input natural language question may be a variety of input terminal 2, for example, using Microsoft Pinyin character input method to the user on the keyboard, the terminal 2 may be carried out to the user through the microphone voice input, in the case of voice input, the input sound related to the recognition process to determine the semantic, because the prior technology is mature, the present invention will not be described in detail.

然后,该智能化自动客户服务设备l中的智能客服引擎服务器72 将所接收到的自然语言通过人机交互技术转换成供计算机识别的信息;再根据所述的供计算机识别的信息输出反馈信息,并将该反馈信息通过所述的互联网WEB服务器73同样以即时通信工具MSN经由互联网发送到用户终端2。 Then, the intelligent automatic customer service equipment l Intelligent Service Engine server 72 stores the received information into a natural language for the computer identified by the human-computer interaction; then the output of the feedback information according to the information for identifying the computer's and the feedback information 73 to the same MSN instant Messenger to the user terminal 2 via the Internet WEB server via the Internet.

具体地,输出反馈信息时,是由智能客服引擎服务器72对输入的自然语言进行语义分析并根据已建立的知识库服务器71中的知识库对其进行语义匹配,以获得预设于该知识库中和所输入自然语言的语义相匹配的答案并将其输出。 Specifically, when the output feedback information 72 of the natural language input by the intelligent call semantic analysis engine server and its semantic matching in accordance with the established knowledge base server 71 in the knowledge base, the knowledge base in order to obtain a predetermined and entered in natural language semantics answer matches and outputs. 如果所述智能客服引擎服务器72没有获得和所输入的自然语言的语义相匹配的答案,则其将判断是否存在和所输入的自然语言的语义相匹配的询问信息,如果存在相应的和所输入的自然语言的语义相匹配的询问信息,则输出该询问信息,向用户提问;然后该智能客服引擎服务器72再次接收用户根据所述询问信息反馈输入的自然语言;接着对反馈输入的自然语言进行语义分析并进行语义匹配;重复上述步骤,直至用户终止在用户终端2上输入自然语言,或者直至该用户从用户终端获得和用户所输入的自然语言相匹配的答案。 If the Smart Service engine server 72 is not obtained and the semantics of the natural language input matches the answer, it is determined whether the inquiry information and the natural language input semantics matches exists, and if the presence of the corresponding input natural language query semantic information matches, the query message is output, the user questions; then the smart server Service engine 72 receives the user again according to the natural language query feedback input; Next, natural language input is the feedback semantic analysis and semantic matching; repeating the above steps until the user terminates the user terminal 2 on the natural language input, or until the user terminal is obtained and the natural language input by the user matches the user answers.

在以上描述的过程中,所述知识库服务器71中的知识库可以采用并列式的数据表,这种情况下,其分为两个数据表,如图3所示,其中,第一数据表用于存储和基于自然语言的特定问题语义相匹配的预设答案,例如,预设答案al和问题ql相关联,预设答案a2和问题q2相关联等等,这些和用户基于自然语言输入的问题相匹配的答案在数据库中的关系是并列的;第二数据表用于存储和特定自然语言的语义相匹配的预设询问信息,例如,预设询问信息al,和问题ql,相关联,预设询问信息a2 ,和问题q 2 ,相关联等等,这些和用户基于自然语言输入的问题相关联的询问信息在数据库中的关系也是并列的。 In the process described above, the knowledge base in the knowledge base server 71 may be employed tile data table, in this case, which is divided into two data tables, shown in Figure 3, wherein the first data table for storing natural language and semantics of the particular problems that match preset answers, e.g., default questions and answers al associated ql, q2 preset answers a2 and associated problems, etc., and these natural language input based on the user answer questions matched relation in the database are juxtaposed; preset query semantic information of the second data table for storing natural language and a particular match, for example, Al preset interrogation information, and issues QL, associated, default asking for information a2, and question 2, q, etc. associated with these issues and ask for information based on user input natural language associated relationships in the database are also parallel.

此外,所述知识库也可以采用树状结构形式的数据表,这种情况下,该数据表中对特定自然语言根据其语义预先分类并逐级排列,如图4所示,该树状数据表是将用户基于自然语言输入的问题逐级地进行分类排列,例如,用户基于自然语言输入的问题经过语义分析并经过语义匹配为Ql类问题时,在该类问题将存在相应的若干条询问信息,可继续用询问或提示的方式使用户进一步回答,例如分为Qll、 Q12、 Q13等问题对用户进行提问,从而在最后得到和用户基于自然语言输入的问题语义相关联的答案。 Furthermore, the knowledge base may be employed in the form of a tree structure of the data table, in this case, the data table previously classified according to its specific natural language semantics stepwise arrangement shown in Figure 4, the tree data the table is arranged to classify the user step by step based on the natural language input, for example, a user-based natural language input and passes through the semantic analysis of Ql based semantic matching problem, there will be such problems in several pieces of the respective interrogation information, so that users can continue to further answer, for example, is divided into Qll, Q12, Q13 and other issues the user with a way to ask questions or tips, so that in the final and get the answer based on user input natural language semantics associated with the problem. 当然,这种树状结构的数据表也可以在用户输入的基于自然语言的问题具有直接相关联的答案时直接得到答案。 Of course the answer directly, data tables Such a structure can also answer questions directly associated natural language-based user input.

为了使所述智能化自动客户服务设备l对用户的回答趋于完善, 知识库服务器71还可存储所接收的自然语言输入内容及其语义匹配结果以更新其中的知识库所包含的例如上述的数据表,所述语义匹配结果包括和用户在用户终端2上输入的自然语言的语义相匹配的答案或者询问信息。 In order to service the client device l intelligent automatic answer user tend to improve, knowledge base server 71 also stores the received content and the natural language input semantic matching result update such as described above wherein the knowledge base contained data table, and the semantic matching result including the natural language semantics of a user input on the user terminal 2 matches the query answers or information.

由于用户在用户终端2上输入的问题多种多样,例如,其中一部分和该政府部门的咨询业务没有直接关系,或者虽然和该政府部分的咨询业务有关但是相应的问题在知识库服务器71的知识库中尚未有相应的或者相关的答案,这种情况下,将不能获得与所接收的自然语言的语义相匹配的答案,甚至也无法根据所接收的自然语言的语义向用户反馈询问信息。 Due to a variety of user input terminal of the user in question 2, for example, where the consulting business and a part of the government departments not directly related to, or knowledge though, and consulting services related to the part of the government but the corresponding problem in the repository server 71 the library has not been a corresponding or relevant answers, in this case, will not be awarded and received the semantics of natural language matches the answer, not even asking for information feedback to the user based on the received semantic natural language. 根据本发明的优选方案,当智能客服引擎服务器 According to a preferred embodiment of the present invention, when the smart customer engine server

72没有获得与所接收的自然语言的语义相匹配的答案和询问信息时, 其可以执行预定的缺省指令,向用户终端2反馈代偿性的内容,例如, 智能客服引擎服务器72可以根据预先设定的缺省信息输出反馈信息; 或者,将所接收的无法获得语义匹配结果的自然语言输入内容通过电子邮件发送至预先设定的邮件地址,该预先设定的邮件地址例如是由专门的人工客服人员进行定时管理;还可以将人机交互界面链接至预先设定的地址,具体方式包括,通过即时通信工具MSN向客户显示一链接地址,引导客户链接到预先设定地址,该预先设定地址可以是由人工客服人员负责的客服电话,或者也可以是表现形式为MSN, QQ 等即时通信工具的人工客服,由人工客服人员补充性地回答所述智能客服引擎服务器72不能获得与所接收的自然语言的语义相匹配的答案的部分问题 72 no answers and semantic inquiry information received matches the natural language, which may be performed a predetermined default commands, feedback to the user terminal 2 compensatory content, e.g., a smart call server engine 72 may advance the default setting information output feedback information; alternatively, the received results can not be obtained semantic matching input natural language content to the e-mail address preset by e-mail, the mail address set in advance, for example, by a dedicated live person management timing; may also be linked to the man-machine interface address set in advance, including the specific embodiment, MSN instant Messenger by displaying a link address to the client, the guide link to the client address set in advance, the preset the address can be set manually in charge of customer service customer service phone, or it may be artificial form of customer service MSN, QQ and other instant messaging tools, artificial customer service supplementary answer by the smart customer service engine server 72 can not be obtained and the part of the problem semantics of natural language matches the received answers

为了使本发明的智能化自动客户服务设备1在使用过程中不断完善和增加知识库服务器71中所存储的知识库内容,在智能化自动客户服务设备1与用户终端进行信息交流的过程中,特别是其中有人工客服人员参与回答智能客服引擎服务器72不能获得与所接收的自然语言的语义相匹配的答案的问题时,知识库服务器71可以存储相关内容,用于更新知识库,例如其可以存储所接收的自然语言输入内容, 以及该自然语言输入内容的语义分析和语义匹配结果,以及其出现次数;以及所接收的无法获得语义匹配的答案和语义匹配的询问信息的自然语言输入内容及其由人工客服人员回复的相应反馈信息;另外, 还包括存储一些涉及专业领域的语言词汇。 In order to make the present invention intelligent automated customer service device 1 during use of continuous improvement and increased knowledge repository content stored in the server 71, the service user terminal device 1 and the process of information exchange in the intelligent automatic client, especially where there is a live person answers smart Service engine participating server 72 issues can not be obtained and the received natural language semantics matches the answer, knowledge base server 71 may store content for updating the knowledge base, which may e.g. natural language input is stored received, and semantic analysis of the natural language input content and semantic matching results, as well as the number of occurrences; and natural language input content inquiry information received can not get semantic matching answers and semantic matching and its reply by a live person corresponding to the feedback information; additionally, further comprising storing a number relates to the field of professional language vocabulary. 通常情况下,知识库服务器71所存储的一项或多项内容任选地可以通过覆盖或者追加方式更新所述知识库月l务器71中的知识库。 Typically, a knowledge base 71 stored in the server or several of the knowledge base can be optionally months in 71 l Service Knowledge updated by overwriting or appended to it.

因此,根据本发明的一个实施方式的优选方案,所述用于构建客服机器人平台的服务器系统实际上包括两个系统:客服机器人管理系 Thus the server system, in accordance with a preferred embodiment of the present embodiment of the invention, the robot platform for constructing the customer actually comprises two systems: Robot Service Management

23统和客服机器人智能引擎系统。 23 robotic systems and customer intelligence engine system.

如图8所示,根据上述优选方案的服务器系统包括:知识库服务器71、智能客服引擎服务器72、以及互联网WEB服务器73。 8, according to the preferred embodiment of the server system comprising: a knowledge base server 71, the customer service intelligent engine server 72, and Internet WEB server 73. 分别由三台服务器执行知识库存储、智能客户服务以及和互联网进行数据交换的功能。 Run the knowledge base storage, intelligent customer service as well as by the three servers and Internet data exchange function. 这些服务器主要用于所述的构成客服机器人智能引擎系统, 它们之间通过线缆或者内部网互连,彼此之间可以进行相互之间的通信,任意两个服务器之间均可以相互交换数据。 The main server configuration for the customer intelligence engine robot system, interconnected by a cable or an intranet, may be in communication with each other between each other therebetween, data can be interchanged between any two servers. 此外,在上述各服务器所构成的客服机器人智能引擎系统的基础上,完整的服务器系统还包括一智能客服管理后台75,其主要用于构成客服机器人管理系统。 In addition, on the basis of customer intelligence engine robot system composed of the above on the server, complete server system also includes a smart customer service management background 75, which constitutes the main customer for robot control system.

所述客服机器人管理系统能够帮助用户管理和部署智能化自动客户服务设备l,。 The robot customer service management system to help users manage and deploy intelligent automated customer service equipment l ,. 它支持UNIX、 Linux、 Windows等操作系统,管理人员可以通过基于WEB的管理后台方便地管理客服系统的内容,如调整客服系统的表现形式等操作。 It supports UNIX, Linux, Windows and other operating systems, administrators can easily manage content-based customer service system WEB management background, such as adjusting the form of customer service systems and other operations. 具体包括:1)、客服知识库的增加、 删除和修改;2)、客服知识库的批量导入,例如本发明采用的操作系统可支持多种方式的数据导入(Excel、 Access、文本文件等),并且可以选择是覆盖/追加等选择项;3)、知识库的处理策略,当数据库中没有与用户的提问相关的问题时,使用何种策略应对,可选择缺省回答、Email给人工客服、转接人工客服(MSN机器人支持)等多种方式; 4)、管理(增加和修改)机器人的欢迎语;5)、管理(增加和修改) 机器人的缺省回答;6)、修改人工客服的信息;7)、日志分析功能, 可以查看系统的详细使用情况,具体功能如下:问题内容,回答内容, IP地址,提问时间,响应时间;针对某IP察看该IP才是问列表以及^J答内容;通过日志分析,系统管理员能标记错误回答,使机器人再次遇到相同问题时可以规避错误答案;提供问题被 Including: 1), increasing customer knowledge base, deletions and modifications; 2), bulk import customer knowledge base, such as an operating system of the present invention may be employed in a variety of ways to support data import (Excel, Access, text files, etc.) and can choose to cover / append and other options; 3), processing strategy knowledge base, when the database is not a problem related to the user's question time, what strategy to use to deal with, you can choose the default answer, Email to a human customer service adapter artificial customer service (MSN robot support) and other methods; 4), management (additions and modifications) robot greeting; 5), default management (additions and modifications) of the robot answered; 6), modified artificial customer service information; 7), detailed usage log analysis function, you can view the system, specific functions are as follows: content, answer content, IP address, question time, response time; is asking for a list of IP to view the IP and ^ J A content; you can avoid the wrong answer when the log analysis, system administrators can mark the wrong answer, the robot once again encountering the same problem; the problem is to provide 问次数的排名;提供筛选功能,能筛选出正确回答、错误回答、供选择的回答、无法回答等情况。 Q. ranked number; provide filtering capabilities, can filter out the right answer, wrong answer, alternative answer, could not answer and so on. 同时提供提问时间筛选、IP地址筛选、问题内容筛选;8)、 分析机器人无法回复的问题列表,添加正确的问题和答案到知识库中,不断的丰富知识库,从而提高机器人的智能;9)、增加专业词汇,丰富机器人的专业领域知识;IO)灵活的管理人员权限设置。 Question time while providing filtering, IP address filtering, content filtering problem; 8), analyze the robot can not reply to a list of questions, the right questions and answers add to the knowledge base, and continue to enrich the knowledge base, thereby enhancing the intelligence of robots; 9) , increase vocabulary, rich domain expertise of the robot; IO) flexible management of permissions. 在利用根据本发明的智能化自动客户服务设备1或7,与用户进行信息交流时,以实现智能化自动客户服务的方法的流程同样可以参照图2,以下结合图7或图8所示的客服机器人平台体系架构示意图以及图2,对所述流程中的具体步骤说明如下: When the service using apparatus 1 or 7, to exchange information with the intelligent user according to the present invention automatically customer, the customer service in order to achieve an automatic method of intelligent processes can also refer to FIG. 2, shown below in conjunction with FIG. 7 or FIG. 8 Service robot platform architecture schematic and 2, the specific steps in the process described below:

步骤S210,用户通过用户终端2以互联网WEB交互方式或者即时通信工具向图2所示的服务器系统构建的客服机器人平台进行信息输入,用户终端2可以将用户以各种输入法输入的文字,或者用户以麦克风输入的声音等数据通过互联网发送给所述客服机器人平台中的互联网WEB服务器73。 Step S210, the user terminal 2 to the Internet WEB or interactively input information to the instant messaging tools to build the server system shown in FIG. 2 by a customer service robot platform user, the user terminal 2 text input by the user can input a variety of methods, or Service user sends to the robot platform internet WEB server 73 through the internet to the microphone input sound data and the like.

步骤S220,互联网WEB服务器73将接收到的数据通过线缆或者局域网发送到智能客服引擎服务器72,由该智能客服引擎服务器72对所输入的自然语言进行语义分析并与预存知识库服务器71中的知识库进行语义匹配,例如,提取所述自然语言中的语义关键词,将其排练组合成该智能客服引擎适于识别或分类检索的语句片段。 Step S220, the Internet WEB server 73 transmits the received data through a cable or LAN to the intelligent call server engine 72, the customer service server engine 72 of the smart semantic analysis of the natural language input and is stored in the knowledge base server 71 matching semantic knowledge base, e.g., natural language to extract the semantic key, the combination thereof into a rehearsal call engine adapted to the smart statement retrieved segment identification or classification. 然后在所述知识库中对其进行语义匹配,以获得和该自然语言的语义相匹配的答案。 Then in the knowledge database semantic matching them to obtain and semantics of the natural language answer matches.

然后执行判断步骤S230,如果在上述步骤之后智能客服引擎服务器72已在所述知识库服务器71中获得和用户输入的自然语言的语义相匹配的答案,则直接执行步骤S260,智能客服引擎服务器72以所述的和用户输入的自然语言的语义相匹配的答案为反馈信息,将其发送至互联网WEB服务器73,并由该互联网WEB服务器73通过互联网以互联网WEB交互方式或者即时通信工具的方式将所述反馈信息发送至用户终端2,即可完成从接收用户输入信息到用户获得反馈信息的信息交流过程。 Then performs determination step S230, the engine if the smart customer server 72 and the user has obtained the natural language input in the knowledge base server 71 after the above-described step matches a semantic answer yes, executing step S260, the server 72 Smart Service Engine according to the semantics of the natural language and user input matches the answer to the feedback information, transmits it to the Internet server 73 WEB by the WEB server 73 via the Internet to the Internet WEB Internet IM tool or interactively manner the feedback information transmitted to the user terminal 2, to complete the input information received from the user to obtain the user feedback information process of information exchange.

如果在上述步骤之后智能客服引擎服务器72在所述知识库服务器71的知识库中不能获得和用户输入的自然语言的语义相匹配的答案,则继续执行判断步骤S240,判断在所述知识库服务器71的知识库中是否存在和用户输入的自然语言的语义相匹配的询问信息。 If the Smart Service Engine and the server 72 can not obtain a user input semantics of the natural language answer matches the knowledge of knowledge base server 71 after the above step, it proceeds to determination step S240, the determination in the knowledge base server whether there is a natural language query and semantic information input by the user matches the knowledge base 71. 如果存在和用户输入的自然语言的语义相匹配的询问信息,则执行步骤 If the query semantic information and the natural language input by the user match exists, execute step

S241,由智能客服引擎服务器72从知识库服务器71中提取出该询问信 S241, 72 extracted by the intelligent call from the knowledge base server engine server 71 that the interrogation channel

其中,所述询问信息是用于向用户提问,例如,当用户输入的自然语言是"我想了解天气情况。"由于天气情况是根据地区和时间而变化, 智能客服引擎服务器72在所述知识库服务器71的知识库中不能获得和用户输入的自然语言的语义相匹配的答案。 Wherein the information is used to ask the user questions, for example, when a user inputs a natural language is "I want to know the weather conditions." Varies depending on weather conditions is due to the region and time, intelligent customer service in the knowledge engine server 72 Knowledge library server 71 can not get natural language and semantic entered by the user matches the answer. 通常,知识库服务器71 的知识库中会存在和该用户输入的问题的语义相关的询问信息"您想知道什么地区,什么时间的天气情况?"这种情况下,客服引擎服务器72就可以在所述知识库中获取该询问信息,然后客服引擎服务器72 Semantically related information is often asked, knowledge base repository server 71 will be problematic and the user input "Would you like to know what area, what the weather conditions of the time?" In this case, the customer service engine server 72 will be in the knowledge acquired in the inquiry message, the server engine 72 and the customer service

用户可以根据用户终端2显示的所述询问信息进一步以自然语言输入信息,例如,"我想知道上海市今天的天气情况。"该输入的自然语言数据将由用户终端通过互联网发送给所述客服机器人平台中的互联网WEB服务器73,重复进行上述从步骤S210开始的步骤。 Users can ask according to the user terminal 2 displays information in a natural language input further information, such as, "I want to know the weather conditions today in Shanghai." Natural language data input by the user of the terminal is sent to the customer service robot via the Internet internet WEB server platform 73, starting from step S210 above is repeated.

如果在上述步骤之后智能客服引擎服务器72在所述知识库服务器71的知识库中不能获得和用户输入的自然语言的语义相匹配的询问信息,所述智能客服引擎服务器7 2将继续执行预定的缺省步骤S250,例如,智能客服引擎服务器72可以根据预先设定的缺省信息输出反馈信息;或者,将所接收的无法获得语义匹配结果的自然语言输入内容通过电子邮件发送至预先设定的邮件地址,该预先设定的邮件地址例如是由专门的人工客服人员进行定时管理;还可以将人机交互界面链接至预先设定的地址,具体方式包括,通过即时通信工具MSN 向客户显示一链接地址,引导客户链接到预先设定地址,该预先设定地址可以是由人工客服人员负责的客服电话,或者也可以是表现形式为MSN, QQ等即时通信工具的人工客服,由人工客服人员补充性地回答所述智能客服引擎服务器72不 If the Smart Service Engine server can not be obtained after the above step 72 asking the user input information and the natural language semantics matches in the knowledge base server repository 71, the intelligent call engine server 72 will continue predetermined default step S250, the example, the server engine 72 may be a smart customer feedback information output from the default preset information; or the received matching result can not be obtained the semantic content of the natural language input by sending mail to a preset e-mail address, the mail address set in advance, for example, is managed by a timer dedicated to a live person; man-machine interface is also linked to the address set in advance, the specific embodiment can be included, by displaying to the customer an MSN instant Messenger link address to guide customers to link to a pre-set address, the pre-set telephone customer service address can be undertaken by a live person, or may be a form of artificial customer service MSN, QQ and other instant messaging tools, customer service staff by hand supplementary answer the call intelligent engine server 72 does not 获得与所接收的自然语言的语义相匹配的答案的部分问题。 Part of the problem is obtained with the semantics of the natural language matches the received answers.

通常情况下,当用户终止输入自然语言或获得和用户所输入的自然语言相匹配的答案后,所述的流程结束。 Typically, when the user terminal input natural language and natural language or obtaining user input matches the answer, the process ends.

结合以上实施例的具体描述,对本发明所提供方法以及实现该方 Described above with reference to specific embodiments, the present invention provides a method and implementing the party

法的设备的优点分析如下: Method advantage of the device is as follows:

1) 降低企业在客户服务中心的再投资,从容应对用户数的大量 1) reduce re-invested enterprises in the customer service center, calmly deal with large amounts of users

增长; increase;

根据对银行、电信、政府和普通企业客户服务中心数据的调查, 客户中心所接到的问题之中很大部分是重复的问题或者简单咨询的 According to a survey of the problem among the banking, telecommunications, government and general corporate customer service center data, customer center received a large part of the problem is repeated or simple advice

问题,例如产品的功能、特点,办事流程等。 Issues, such as product features, characteristics, work processes and so on. 如果我们可以将这部分问题分流到其他渠道,例如企业的网站将能够满足这部分用户的需求。 If we can bypass this part of the problem to other channels, such as corporate Web site will be able to meet this part of the user's needs. 企业通常都试图这么做,不过根据最终用户的反馈,他们在访问企业的网站时通常都觉得网站结构太过复杂,需要花费太多的时间来寻找自己需要的内容。 Businesses often are trying to do, but according to the end-user feedback when they visit company websites usually feel the site structure is too complex, take too much time to find the content they need.

本发明的客服机器人或者智能化自动客户服务设备能够通过引导性的问答,了解客户的需求,在较短的时间内给用户一个满意的回答。 Service robots invention or device capable of intelligent automated customer service by question and answer introductory, understand customer needs, give the user a satisfactory answer in a short period of time. 且用户只需要像和客服代表一样聊天就能够达到自己的目的。 And the user only needs to like and chat as a customer service representative will be able to achieve their own ends. 且由于系统架构本身的特点,本发明能够实现同时对大量用户的服务。 And because of the characteristics of the system architecture itself, the present invention enables a large number of simultaneous users of the service.

2) 降低企业客服中心的运行成本; 2) reduce operating costs of call center;

根据Forrester Research研究显示,有30%的顾客会透过网络上的Self Service方式来获得答案。 According to Forrester Research study showed that 30% of customers will come to get answers through the Self Service mode on the network. 根据Gamer Group研究显示,客服回复电话的成本为$5.50/个,而使用网络客服的成本仅为$0.24/个;通过将部分咨询问题转到成本更低的网络客服,企业可以节省更多的通信费用和人员成本。 According to Gamer Group research shows that the cost of customer service return calls for $ 5.50 / month, while the cost of using the network of customer service is only $ 0.24 / month; ask questions through the part to lower the cost of customer service network, companies can save more communication costs and personnel costs.

3) 减少客户的不愉快体验-等待时间、回答的准确性 3) reduce customer unpleasant experience - the waiting time, the accuracy of answers

由于将部分咨询问题转移到网络客服,企业有更多的更多的资源(线路、人员)投入到复杂的问题之中。 Due to the transfer of part of the problem to consult the customer service network, enterprises have more more resources (line, staff) put into complex issues. 采用本发明的技术方案,对于企业而言,其每次拨打客服中心时,需要排队的时间就更短,甚至能够即时接通。 The technical solution of the present invention, for enterprises, which call the customer service center every time, even shorter queuing time required, even immediate access.

而对于使用网络客服的用户来说,系统能够根据其问题即时给出相关答复,而且由于其回答基于企业标准信息,不会存在错误。 For users using the network of customer service, the system can immediately give its reply in accordance with related problems, and because of its standards-based enterprise information answer, there will be no errors. 5) 对客户进行分析并自我更新服务能力; 系统能够完整记录所有的用户问题,企业或者政府能够根据这些 5) analysis of the customer service and self-renewal capacity; the system can complete record of all customer issues, according to these companies or government can

记录分析客户的行为特点,例如哪些问题是用户最关心的,特定用户对什么产品/服务感兴趣等。 Recording and analyzing behavioral characteristics of customers, such as which users are most concerned about is the issue of the specific user is interested in what products / services. 而利用传统客户,这些信息保存于语音形式或者属于客服人员的私有信息,很难进行信息的再加工(搜索和分析)。 The use of traditional customers, such information is stored in the form of voice or proprietary information belonging to customer service, it is difficult for reprocessing information (search and analysis).

系统通过对常见问题的分析,可以提供更为准确的回答。 System through the analysis of common problems, may provide a more accurate answer. 对于暂时无法满足的问题,企业可以进行分析,从而为以后类似的问题提供信息。 For the time being the problem can not be met, companies can be analyzed to provide information on similar problems in the future.

6) 低成本提供昼夜和假日服务 6) provides low-cost day and night and holiday service

在假日和下班时段内,用户愿意使用更长的时间解决产品或者服务的问题,在传统的客服务中心模式下,企业或者政府一方面要雇用更多的客服代表,另一方面平均通话时间变长,相应的成本会急剧增加。 And work in the holiday period, users are willing to use a longer time to solve the problem of product or service, in the traditional customer service center model, the enterprise or the Government should hire more customer service representatives, on the other hand the average talk time change long, the corresponding costs would increase drastically. 而贏思智能客服机器人一旦部署,能够在几乎无成本增加的情况下提供全时服务。 The Incesoft intelligent customer service robot once deployed, can provide full-time service with almost no increase in the cost of the case.

客服机器人作为一种全新的客服工具,主要用于解决人工客服不能解决的问题,其投入成本低,应用效果好,可以24小时在线实时回复用户提问。 Service robots as a new customer service tool, mainly used to solve the problem can not be solved artificial customer service, low input costs, good application effect, 24-hour online live reply user questions. 应用这项技术的单位,其为公众服务的能力得到明显提高。 The application of this technology unit, its ability to serve the public has been significantly improved. 客服机器人能够帮助所述单位的各级部门方便的从人工客服方式过渡到电脑全自动人工智能型的机器人客服方式。 Service robots can help the units at all levels to facilitate the transition from manual to automatic computer customer service manner of artificial intelligence robot customer service way. 通过网站,可以完整的传递信息,结合图片、文字甚至视频给用户最完整的回复,让用户在同客服机器人愉快轻松的交流中解决问题。 Through the website, you can complete the transfer of information, with pictures, text and even video to users the most complete response to let users solve problems with customer service robot in a pleasant relaxed communication.

总之,企业或者政府等单位通过部署智能客服机器人或者智能化自动客户服务设备,能够为传统客户服务中心提供问题分流支持,使传统服务中心能够专注于复杂的和高价值的客户。 In short, companies or government and other units of intelligent customer service robot or automated customer service by deploying intelligent devices, to provide for the issue of traditional customer service center to support the shunt, the traditional service centers to focus on complex and high-value customers. 从而在两种渠道内,企业能为用户提供了更为迅速、准确而标准的服务。 So in the two channels, companies can provide a more rapid, accurate and standard of services for users. 在低成本的条件下,为企业的高速发展提供了有效的支持。 Under conditions of low-cost, high-speed development of enterprises to provide effective support. 对于企业或者政府来说,这种领先技术的使用也为其增加了新的形象(科技和创新的形象)。 For business or government, the use of this leading-edge technology but also for its added a new image (image technology and innovation). 以上对本发明的实施例进行了描述,但是本发明并不局限于特定的系统、设备和具体协议,本领域内技术人员可以在所附权利要求的范围内做出各种变形或修改。 For the above embodiments of the present invention has been described, but the present invention is not limited to specific systems, specific devices and protocols, those skilled in the art can make various changes and modifications within the scope of the appended claims.

Claims (16)

1. 一种通过人机交互技术实现自动客户服务的方法,其特征在于,包括以下步骤:a接收用户输入的基于自然语言的查询信息;b根据所述基于自然语言的查询信息输出查询结果。 A method for implementing automated customer service through human-computer interaction, which is characterized in that it comprises the steps of: a user input receiving a query based on natural language; b. A query result based on the query information based on an output of the natural language.
2. 根据权利要求l所述的方法,其特征在于,所述步骤a包括:-接收用户通过文字或者语音输入的基于自然语言的查询信息。 2. The method according to claim l, wherein said step a comprises: - receiving a user query information by voice or text-based natural language input.
3.根据权利要求1或2所述的方法,其特征在于,所述步骤b包括: bl对作为用户输入信息的查询信息进行语义分析,并与所述预存的知识库中的内容进行语义匹配,获得与该查询信息相匹配的问题,所述知识库包含具有一定拓朴结构的多个问题及其相关联的答案;b2当所述语义匹配获得的问题具有对应的答案,将该答案作为查询结果输出。 3. The method of claim 1 or claim 2, wherein said step b comprises: bl query information as the user information input semantic analysis, and semantic matching with the content stored in the Knowledge Base , the problem of obtaining query information matches, said knowledge base comprising a plurality of questions having a certain topology and associated answers; B2 matches the semantic problems when having obtained a corresponding answer, and the answers as query results output.
4. 根据权利要求3所述的方法,其特征在于,所述步骤b还包括下列步骤:b3当所述语义匹配获得的问题没有对应的答案而仅具有相关联的询问信息,则将该询问信息输出;b4接收用户针对所述询问信息而输入的基于自然语言的反馈信自.将所述反馈信息作为所述用户输入信息来重复上述步骤bl至b4, 直至获得所述对应的答案和所述反馈信息为用户指示结束查询的指示信息。 4. The method according to claim 3, wherein said step b further comprises the steps of: b3 matching the semantic problems when no corresponding answer obtained only information associated with the interrogation, the interrogation is information output;. b4 receiving user information for the query based on the natural language input feedback signal from the feedback information as the user input information to repeat the above steps bl to b4, until the answers and the corresponding said feedback indication information indicating the end of the user query.
5. 根据权利要求3或4所述的方法,其特征在于,还包括下列步骤: -存储所述用户输入的基于自然语言的查询信息和/或反馈信息及其语义匹配结果以更新所述知识库,所述语义匹配结果包括所述语义匹配的问题及所对应的答案或者询问信息。 The method according to claim 3 or claim 4, characterized by further comprising the steps of: - natural language based query information and / or feedback information to update the matching results and Semantic storing the knowledge of the user input library, comprising the semantic matching result matches the semantic problems and answers corresponding to information or ask.
6. 根据权利要求3或4所述的方法,其特征在于,当没有获得与所接收的自然语言的语义相匹配的答案和询问信息时,执行以下各步骤中的任一个或任多个:-根据预先设定的缺省信息输出反馈信息;-将所接收的无法获得语义匹配结果的自然语言输入内容通过电子邮件发送至预先设定的邮件地址;-将所述人机交互界面链接至预先设定的地址。 The method according to claim 3 or claim 4, wherein, when no answer is obtained, and semantic information inquiry received natural language matches any of the one or more of any of performing the following steps: - the output of the feedback information based on the default preset information; - the received natural language input can not be obtained semantic content of the matching result via e-mail message sent to a predetermined address; - the interactive interface linked to a pre-set address.
7. 根据权利要求6所述的方法,其特征在于,存储下列的任一项或任多项内容,用于更新知识库:-所接收的所述用户输入的基于自然语言的查询信息和/或反馈信息及其语义匹配结果,以及其出现次数;-所接收的所述知识库中无法获得语义匹配的答案和语义匹配的询问信息的自然语言输入内容及其相应的反馈信息;-涉及专业领域的语言词汇。 7. The method according to claim 6, wherein any one or any number of the following stored contents to update the knowledge base: - query information received by the user based on the input natural language and / or feedback and semantic matching results, and their number of occurrences; - the knowledge of the received inquiry information can not be obtained in the natural language semantic matching answers and input semantic matching content and the corresponding feedback information; - involving professional Vocabulary field.
8. 根据权利要求7所述的方法,其特征在于,所存储的一项或多项内容以替换或者追加方式更新所述知识库。 8. The method according to claim 7, wherein one or more of the stored content to replace or append mode updating the knowledge base.
9. 一种通过人机交互技术实现自动客户服务的设备,其特征在于,包括:人机交互装置,用于接收用户输入的基于自然语言的查询信息; 反馈装置,用于根据所述基于自然语言的查询信息输出查询结果。 9. An apparatus for automatic customer service through human-computer interaction, characterized by comprising: man-machine interactive means for receiving a user input query information based on natural language; feedback means, according to the natural-based query language information output query results.
10. 根据权利要求9所述的设备,其特征在于,所述人机交互装置用于接收用户通过文字或者语音输入的基于自然语言的查询信息。 10. The apparatus according to claim 9, wherein said interactive means for receiving a user query via text or voice information based on the input natural language.
11. 根据权利要求9或10所述的设备,其特征在于,所述反馈装置包括:语义匹配装置,用于对作为用户输入信息的查询信息进行语义分析,并与所述预存的知识库中的内容进行语义匹配,获得与该查询信息相匹配的问题,所述知识库包含具有一定拓朴结构的多个问题及其相关联的答案;当所述语义匹配获得的问题具有对应的答案,该语义匹配装置将该答案作为查询结果输出。 11. The apparatus according to claim 9 or claim 10, wherein said feedback means comprises: knowledge base semantic matchmaking device, the query information for the user input information as a semantic analysis, with said pre-stored in semantic content matching, problems with obtaining information that matches the query, the knowledge base comprising a plurality of questions having a certain topology and associated answers; matching the semantic problems when having obtained a corresponding answer, the semantic matchmaking device answers the query result as an output.
12. 根据权利要求ll所述的设备,其特征在于,所述语义匹配装置还用于:-当所述语义匹配获得的问题没有对应的答案而仅具有相关联的询问信息,则将该询问信息输出;-接收用户针对所述询问信息而输入的基于自然语言的反馈信自、./:、 5 12. The apparatus as claimed in claim ll, wherein, the semantic matchmaking device is further configured to: - the semantic matching is obtained when the problem is not only the answers corresponding to information associated with query, then the query information output; - receiving a feedback signal for the user query based on the input information from natural language, / :, 5.
13. 根据权利要求11或12所述的设备,其特征在于,所述反馈装置还包括:知识库存储装置,用于存储所述用户输入的基于自然语言的查询信息和/或反馈信息及其语义匹配结果以更新所述知识库,所述语义 13. The apparatus of claim 11 or claim 12, wherein said feedback means further comprising: a knowledge base storage means for storing the user input information based on natural language queries and / or feedback information and the semantic matching results to update the knowledge base, the semantic
14. 根据权利要求11或12所述的设备,其特征在于,当没有获得与所接收的自然语言的语义相匹配的答案和询问信息时,所述反馈装置执行以下各步骤中的任一个或任多个:-根据预先设定的缺省信息输出反馈信息;-将所接收的无法获得语义匹配结果的自然语言输入内容通过电子邮件发送至预先设定的邮件地址;-将所述人机交互界面链接至预先设定的地址。 14. The apparatus of claim 11 or claim 12, wherein, when no answer is obtained, and semantic information inquiry received natural language matches any of the following steps or means for performing feedback any one of a plurality of: - the output of the feedback information based on the default information set in advance; - the received natural language input can not be obtained semantic content of the matching result via e-mail message sent to a predetermined address; - the said human interactive interface linked to the address set in advance.
15. 根据权利要求14所述的设备,其特征在于,所述知识库存储装置用于存储下列的任一项或任多项内容,以更新知识库:-所接收的所述用户输入的基于自然语言的查询信息和/或反馈信息及其语义匹配结果,以及其出现次数;-所接收的所述知识库中无法获得语义匹配的答案和语义匹配的询问信息的自然语言输入内容及其相应的反馈信息;-涉及专业领域的语言词汇。 15. The apparatus according to claim 14, wherein said knowledge base storage means for storing any one or any of the following more content to update the knowledge base: - based on the received user input natural language query information and / or feedback information and semantic matching result, and the frequency of its occurrence; - the received knowledge can not be obtained in the natural language query semantic matching answer information and semantic content and the corresponding input matching feedback; - relates to the field of professional language vocabulary.
16. 根据权利要求15所述的设备,其特征在于,所述知识库存储装置所存储的一项或多项内容以覆盖或者追加方式更新所述知识库。 16. Apparatus according to claim 15, wherein the one or more knowledge base stored in the content storage means so as to cover the knowledge base to update or append mode.
CN 200710048039 2007-11-08 2007-11-08 Method and equipment for implementing automatic customer service through human-machine interaction technology CN101431573B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN 200710048039 CN101431573B (en) 2007-11-08 2007-11-08 Method and equipment for implementing automatic customer service through human-machine interaction technology

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN 200710048039 CN101431573B (en) 2007-11-08 2007-11-08 Method and equipment for implementing automatic customer service through human-machine interaction technology
HK09105532A HK1128110A1 (en) 2007-11-08 2009-06-19 Method and apparatus for providing automatic customer services through man- computer interaction

Publications (2)

Publication Number Publication Date
CN101431573A true true CN101431573A (en) 2009-05-13
CN101431573B CN101431573B (en) 2013-02-20

Family

ID=40646722

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 200710048039 CN101431573B (en) 2007-11-08 2007-11-08 Method and equipment for implementing automatic customer service through human-machine interaction technology

Country Status (1)

Country Link
CN (1) CN101431573B (en)

Cited By (29)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102622369A (en) * 2011-01-26 2012-08-01 吴凯 Novel efficient real-time network information platform system with human thinking characteristics
CN102624675A (en) * 2011-01-27 2012-08-01 腾讯科技(深圳)有限公司 Self-service customer service system and method
CN102739557A (en) * 2011-04-08 2012-10-17 腾讯科技(深圳)有限公司 User distribution method and device in customer services
CN102789496A (en) * 2012-07-13 2012-11-21 携程计算机技术(上海)有限公司 Method and system for implementing intelligent response
CN102868695A (en) * 2012-09-18 2013-01-09 天格科技(杭州)有限公司 Conversation tree-based intelligent online customer service method and system
CN102968987A (en) * 2012-11-19 2013-03-13 百度在线网络技术(北京)有限公司 Speech recognition method and system
CN103294725A (en) * 2012-03-03 2013-09-11 李辉 Intelligent response robot software
CN103297503A (en) * 2013-05-08 2013-09-11 南京邮电大学 Mobile terminal swarm intelligent perception structure based on layered information extraction server
CN103312704A (en) * 2013-06-03 2013-09-18 苏州阔地网络科技有限公司 Method and system for realizing intelligent customer service
CN103365836A (en) * 2012-04-01 2013-10-23 郭佳 Natural language utilized distributed intelligent interaction achieving method and system thereof
CN103795876A (en) * 2012-10-29 2014-05-14 殷程 Intelligent voice script
CN103941865A (en) * 2014-04-08 2014-07-23 三和智控(北京)系统集成有限公司 Method for establishing discrete human-computer interaction mechanism
CN104268667A (en) * 2014-09-03 2015-01-07 陈飞 Method and system for assisting manager in dispute resolution
CN104424352A (en) * 2013-08-22 2015-03-18 乐金信世股份有限公司 System and method for providing agent service to user terminal
CN104469029A (en) * 2014-11-21 2015-03-25 科大讯飞股份有限公司 Method and device for telephone number query through voice
CN104679815A (en) * 2014-12-08 2015-06-03 北京云知声信息技术有限公司 Method and system for screening question and answer pairs and updating question and answer database in real time
CN104735132A (en) * 2015-02-13 2015-06-24 腾讯科技(深圳)有限公司 Information inquiry method, servers and terminal
CN104809197A (en) * 2015-04-24 2015-07-29 同程网络科技股份有限公司 On-line question and answer method based on intelligent robot
CN104951458A (en) * 2014-03-26 2015-09-30 华为技术有限公司 Method and equipment for helping processing based on semantic recognition
CN105072173A (en) * 2015-08-03 2015-11-18 谌志群 Customer service method and system for automatically switching between automatic customer service and artificial customer service
WO2016071917A1 (en) * 2014-11-03 2016-05-12 Hewlett-Packard Development Company, L.P. Social media analytics and response
CN105721725A (en) * 2016-02-03 2016-06-29 北京光年无限科技有限公司 Customer service oriented question and answer interaction method and system
CN105808694A (en) * 2016-03-04 2016-07-27 上海携程商务有限公司 Online customer service response system and method
CN105930367A (en) * 2016-04-12 2016-09-07 华南师范大学 Intelligent chatting robot control method and control device
CN105978986A (en) * 2016-06-13 2016-09-28 镇江市润州园林绿化研究所 Scientific advisory system for garden landscape ecology
CN106078764A (en) * 2016-08-12 2016-11-09 李乾 Elderly companion robot and control method thereof
WO2017032084A1 (en) * 2015-08-24 2017-03-02 北京云知声信息技术有限公司 Information output method and apparatus
CN107688614A (en) * 2017-08-04 2018-02-13 平安科技(深圳)有限公司 Intension acquiring method, electronic device and computer readable storage medium
CN105592234B (en) * 2014-10-21 2018-08-28 阿里巴巴集团控股有限公司 A communication data processing method and apparatus

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105338204A (en) * 2014-08-15 2016-02-17 中兴通讯股份有限公司 Interactive voice response method and device
CN104505091B (en) * 2014-12-26 2018-08-21 湖南华凯文化创意股份有限公司 Human-computer interaction method and system for voice

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1741013A (en) 2004-08-27 2006-03-01 英业达股份有限公司 Automatic adaptive system for customer service and its method
CN1821991B (en) 2005-02-18 2010-04-28 上海赢思软件技术有限公 Knowledge question-and-answer quick processing system based on artificial intelligence
CN100578539C (en) 2006-02-28 2010-01-06 腾讯科技(深圳)有限公司 Automatic question-answering method and system
CN100416570C (en) 2006-09-22 2008-09-03 浙江大学 FAQ based Chinese natural language ask and answer method

Cited By (39)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102622369A (en) * 2011-01-26 2012-08-01 吴凯 Novel efficient real-time network information platform system with human thinking characteristics
CN102624675A (en) * 2011-01-27 2012-08-01 腾讯科技(深圳)有限公司 Self-service customer service system and method
CN102624675B (en) 2011-01-27 2014-08-06 腾讯科技(深圳)有限公司 Self-service customer service system and method
CN102739557B (en) * 2011-04-08 2016-06-22 腾讯科技(深圳)有限公司 Users shunt method and apparatus for customer service
CN102739557A (en) * 2011-04-08 2012-10-17 腾讯科技(深圳)有限公司 User distribution method and device in customer services
CN103294725A (en) * 2012-03-03 2013-09-11 李辉 Intelligent response robot software
CN103365836B (en) * 2012-04-01 2016-05-11 郭佳 Distributed intelligent interaction using a natural language system and its implementation
CN103365836A (en) * 2012-04-01 2013-10-23 郭佳 Natural language utilized distributed intelligent interaction achieving method and system thereof
CN102789496A (en) * 2012-07-13 2012-11-21 携程计算机技术(上海)有限公司 Method and system for implementing intelligent response
CN102868695B (en) * 2012-09-18 2015-06-17 天格科技(杭州)有限公司 Conversation tree-based intelligent online customer service method and system
CN102868695A (en) * 2012-09-18 2013-01-09 天格科技(杭州)有限公司 Conversation tree-based intelligent online customer service method and system
CN103795876A (en) * 2012-10-29 2014-05-14 殷程 Intelligent voice script
CN102968987A (en) * 2012-11-19 2013-03-13 百度在线网络技术(北京)有限公司 Speech recognition method and system
CN103297503A (en) * 2013-05-08 2013-09-11 南京邮电大学 Mobile terminal swarm intelligent perception structure based on layered information extraction server
CN103297503B (en) * 2013-05-08 2016-08-17 南京邮电大学 Chilean mobile terminal group sensing system based on hierarchical information extraction server
CN103312704A (en) * 2013-06-03 2013-09-18 苏州阔地网络科技有限公司 Method and system for realizing intelligent customer service
CN104424352A (en) * 2013-08-22 2015-03-18 乐金信世股份有限公司 System and method for providing agent service to user terminal
CN104424352B (en) * 2013-08-22 2017-10-13 乐金信世股份有限公司 Provision of agency services to the user terminal system and method
WO2015144065A1 (en) * 2014-03-26 2015-10-01 华为技术有限公司 Semantic recognition-based help processing method and device
CN104951458A (en) * 2014-03-26 2015-09-30 华为技术有限公司 Method and equipment for helping processing based on semantic recognition
CN103941865A (en) * 2014-04-08 2014-07-23 三和智控(北京)系统集成有限公司 Method for establishing discrete human-computer interaction mechanism
CN104268667A (en) * 2014-09-03 2015-01-07 陈飞 Method and system for assisting manager in dispute resolution
CN105592234B (en) * 2014-10-21 2018-08-28 阿里巴巴集团控股有限公司 A communication data processing method and apparatus
WO2016071917A1 (en) * 2014-11-03 2016-05-12 Hewlett-Packard Development Company, L.P. Social media analytics and response
CN104469029A (en) * 2014-11-21 2015-03-25 科大讯飞股份有限公司 Method and device for telephone number query through voice
CN104469029B (en) * 2014-11-21 2017-11-07 科大讯飞股份有限公司 Method by voice and the check number, means
CN104679815B (en) * 2014-12-08 2018-02-23 北京云知声信息技术有限公司 A method of screening for Q & A and real-time updates Q & A method and system libraries
CN104679815A (en) * 2014-12-08 2015-06-03 北京云知声信息技术有限公司 Method and system for screening question and answer pairs and updating question and answer database in real time
CN104735132A (en) * 2015-02-13 2015-06-24 腾讯科技(深圳)有限公司 Information inquiry method, servers and terminal
CN104735132B (en) * 2015-02-13 2017-05-24 腾讯科技(深圳)有限公司 Information query method, and terminal server
CN104809197A (en) * 2015-04-24 2015-07-29 同程网络科技股份有限公司 On-line question and answer method based on intelligent robot
CN105072173A (en) * 2015-08-03 2015-11-18 谌志群 Customer service method and system for automatically switching between automatic customer service and artificial customer service
WO2017032084A1 (en) * 2015-08-24 2017-03-02 北京云知声信息技术有限公司 Information output method and apparatus
CN105721725A (en) * 2016-02-03 2016-06-29 北京光年无限科技有限公司 Customer service oriented question and answer interaction method and system
CN105808694A (en) * 2016-03-04 2016-07-27 上海携程商务有限公司 Online customer service response system and method
CN105930367A (en) * 2016-04-12 2016-09-07 华南师范大学 Intelligent chatting robot control method and control device
CN105978986A (en) * 2016-06-13 2016-09-28 镇江市润州园林绿化研究所 Scientific advisory system for garden landscape ecology
CN106078764A (en) * 2016-08-12 2016-11-09 李乾 Elderly companion robot and control method thereof
CN107688614A (en) * 2017-08-04 2018-02-13 平安科技(深圳)有限公司 Intension acquiring method, electronic device and computer readable storage medium

Also Published As

Publication number Publication date Type
CN101431573B (en) 2013-02-20 grant

Similar Documents

Publication Publication Date Title
US7792773B2 (en) Method and system for enabling automated and real-time discovery of skills available to agents and systems in a multimedia communications network
US7216121B2 (en) Search engine facility with automated knowledge retrieval, generation and maintenance
US6889375B1 (en) Method and system for application development
US6973620B2 (en) Method and apparatus for providing user support based on contextual information
Odubiyi et al. SAIRE—a scalable agent-based information retrieval engine
US20060198504A1 (en) Call recording platform
US6381640B1 (en) Method and apparatus for automated personalization and presentation of workload assignments to agents within a multimedia communication center
US6745178B1 (en) Internet based method for facilitating networking among persons with similar interests and for facilitating collaborative searching for information
US20060085417A1 (en) Method and apparatus for data mining within communication session information using an entity relationship model
US20040068424A1 (en) System and method for managing workflow among members of an organization
US20040088300A1 (en) Management system for a contact centre
US20050289124A1 (en) Systems and methods for processing natural language queries
US7460654B1 (en) Processing of enterprise messages integrating voice messaging and data systems
US20060080107A1 (en) Management of conversations
US20080228504A1 (en) Technique to Deflect Incident Ticket Submission in Real-Time
US6170011B1 (en) Method and apparatus for determining and initiating interaction directionality within a multimedia communication center
US20090074159A1 (en) Messaging and application system integration
US6778644B1 (en) Integration of voice messaging and data systems
US20110276598A1 (en) System for and method of providing reusable software service information based on natural language queries
US6346952B1 (en) Method and apparatus for summarizing previous threads in a communication-center chat session
US8346563B1 (en) System and methods for delivering advanced natural language interaction applications
US20050187794A1 (en) Electronic medical record registry including data replication
US20040123242A1 (en) Context instantiated application protocol
US20090190728A1 (en) System and Method for Providing Audible Spoken Name Pronunciations
US20110125550A1 (en) Method for determining customer value and potential from social media and other public data sources

Legal Events

Date Code Title Description
C06 Publication
C10 Entry into substantive examination
REG Reference to a national code

Ref country code: HK

Ref legal event code: DE

Ref document number: 1128110

Country of ref document: HK

COR Change of bibliographic data

Free format text: CORRECT: ADDRESS; FROM: 200000 CHANGNING, SHANGHAI TO: 201700 QINGPU, SHANGHAI

ASS Succession or assignment of patent right

Owner name: SHANGHAI HANSHI NETWORK TECHNOLOGY CO., LTD.

Free format text: FORMER OWNER: SHANGHAI INCESOFT CO., LTD.

Effective date: 20111012

C41 Transfer of patent application or patent right or utility model
C14 Grant of patent or utility model
COR Change of bibliographic data

Free format text: CORRECT: ADDRESS; FROM: 201700 QINGPU, SHANGHAI TO: 200060 PUTUO, SHANGHAI

ASS Succession or assignment of patent right

Owner name: ZHU PINPIN

Free format text: FORMER OWNER: SHANGHAI HANSHI NETWORK TECHNOLOGY CO., LTD.

Effective date: 20130709

C41 Transfer of patent application or patent right or utility model
REG Reference to a national code

Ref country code: HK

Ref legal event code: GR

Ref document number: 1128110

Country of ref document: HK

ASS Succession or assignment of patent right

Owner name: SHANGHAI ZHIZHEN NETWORK TECHNOLOGY CO., LTD.

Free format text: FORMER OWNER: ZHU PINPIN

Effective date: 20141115

COR Change of bibliographic data

Free format text: CORRECT: ADDRESS; FROM: 200060 PUTUO, SHANGHAI TO: 201824 JIADING, SHANGHAI

C41 Transfer of patent application or patent right or utility model
C56 Change in the name or address of the patentee
EE01