WO2015188454A1 - 一种快速进入ivr菜单的方法及装置 - Google Patents
一种快速进入ivr菜单的方法及装置 Download PDFInfo
- Publication number
- WO2015188454A1 WO2015188454A1 PCT/CN2014/084951 CN2014084951W WO2015188454A1 WO 2015188454 A1 WO2015188454 A1 WO 2015188454A1 CN 2014084951 W CN2014084951 W CN 2014084951W WO 2015188454 A1 WO2015188454 A1 WO 2015188454A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- user
- menu
- voice
- ivr
- text
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 36
- 230000008569 process Effects 0.000 claims description 17
- 238000007781 pre-processing Methods 0.000 claims description 6
- 238000001914 filtration Methods 0.000 claims description 3
- 230000002452 interceptive effect Effects 0.000 claims description 3
- 230000004044 response Effects 0.000 claims description 3
- 230000008685 targeting Effects 0.000 claims description 2
- 230000003993 interaction Effects 0.000 abstract description 8
- 230000006870 function Effects 0.000 description 11
- 230000006399 behavior Effects 0.000 description 9
- 230000014509 gene expression Effects 0.000 description 4
- 230000011218 segmentation Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/60—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M3/00—Automatic or semi-automatic exchanges
- H04M3/42—Systems providing special services or facilities to subscribers
- H04M3/50—Centralised arrangements for answering calls; Centralised arrangements for recording messages for absent or busy subscribers ; Centralised arrangements for recording messages
Definitions
- the present invention relates to the field of IVR (Interactive Voice Response) system, and in particular, to a method and apparatus for quickly entering an IVR menu.
- IVR Interactive Voice Response
- the IVR system is an important portal for human-computer interaction in the call center.
- the way the user interacts with the system is mainly through the keyboard of the telephone.
- users often need to press the button multiple times or wait for the menu prompt voice to enter the corresponding IVR menu item, so that the interaction between the user and the IVR system is greatly limited. It has also become the most dissatisfied place in the current user experience.
- the user accidentally presses the wrong button when entering the certain level of the IVR menu, which may result in returning to the main menu. At this time, re-selecting the menu item from top to bottom will waste the user's time.
- Great impact on the efficiency of users using the IVR system is very important portal for human-computer interaction in the call center.
- the technical problem to be solved by the embodiments of the present invention is to provide a method and apparatus for quickly entering an IVR menu, which overcomes the defects caused by entering a IVR menu item by means of a button in the prior art.
- the technical solution adopted by the embodiment of the present invention is that the method for quickly entering the IVR menu includes: providing a menu item matching the user voice to the user according to the user voice matching in the IVR menu.
- the matching based on the user voice in the IVR menu comprises: establishing a menu item index based on the text corresponding to the IVR menu; matching the text corresponding to the user voice with the menu item index, and finding a menu item matching the user voice .
- the IVR menu is an IVR menu that has been flattened; the flattening process includes: for a menu item of two adjacent levels in the IVR menu, the menu item of the first level is corresponding to the next menu item instead.
- the acquiring process of the text corresponding to the user voice includes: After prompting the menu item function that the user can use, the user voice is recorded and converted into user voice text, and the user voice text is preprocessed, and the preprocessing includes: stopping the word processing, and filtering out the non- GBK (Chinese Character Internal Code Extension Specification) Encoded characters and non-text characters.
- the acquiring process of the text corresponding to the user voice further includes: segmenting the pre-processed user voice text.
- the providing, to the user, the menu item that matches the voice of the user includes:
- step A1 determining whether the number of menu items matching the user is 1, if yes, providing the business logic function of the menu item to the user, otherwise performing step A2;
- the user is voice-guided based on the user's voice text, prompting the user to input a more accurate voice and re-matching.
- the embodiment of the invention further provides an apparatus for quickly entering an IVR menu, comprising: a matching module configured to perform matching in an IVR menu based on a user voice; and a processing module configured to provide a menu item matching the user voice to the user.
- the matching module includes: an index establishing module, configured to establish a menu item index based on the text corresponding to the IVR menu; and a text matching module configured to match the text corresponding to the user voice with the menu item index, and find the user Voice-matched menu items.
- the matching module includes: a menu processing module, configured to first flatten the IVR menu and then provide the matching module to the matching module; the flattening process includes: targeting two adjacent levels in the IVR menu The menu item replaces the menu item of the current level with the corresponding next level menu item.
- the present invention also provides an IVR system, comprising: a recording device configured to record a user voice and convert it into a user voice text; The processor is configured to establish a menu item index based on the text corresponding to the IVR menu, match the user voice text with the menu item index, find a menu item that matches the user voice, and provide the service logic function of the menu item to the user.
- the method and device for quickly entering the IVR menu according to the embodiment of the present invention enable the user to quickly reach the menu item in the required IVR menu through natural language interaction, thereby saving The waiting time of the user, and improve the experience of the IVR; preferably, if the IVR menu is flattened, the menu level is reduced; when the user voice input is not accurate, the voice can be actively generated according to the voice input by the user.
- the user performs a step-by-step boot so that the user can quickly enter the desired menu.
- FIG. 1 is a flow chart of a method for quickly entering an IVR menu according to a first embodiment of the present invention
- FIG. 2 is a schematic structural diagram of a device for quickly entering an IVR menu according to a second embodiment of the present invention
- FIG. 4 is a flowchart of execution of a semantic analysis module according to a third embodiment of the present invention
- FIG. 5 is a flow chart of quickly entering an IVR menu according to a third embodiment of the present invention.
- Step S101 Perform matching in an IVR menu based on user voice.
- step S101 includes the following processing flow: A1: acquiring text corresponding to the user voice; further, after prompting the user to use the menu item function, recording the user voice and converting the user voice into the user voice text, the user voice
- the text is preprocessed, and the preprocessing includes: stopping the word processing, And use regular expressions to filter out non-GBK-encoded characters and non-text characters, typical non-text characters, such as: punctuation, and &# characters, to stop words can be based on the stop word dictionary.
- the pre-processed user voice text is the text corresponding to the obtained user voice.
- the pre-processed user voice text is further divided into words, and the user-phone text after the word segmentation is used as the text corresponding to the acquired user voice.
- the word segmentation may be based on word-of-speech or word meaning to decompose the user's phonetic text to better match the keywords in the menu item index.
- A2 establishing a menu item index based on the text corresponding to the IVR menu; preferably, the IVR menu is composed of each menu item. If there are multiple levels of menu items, then a certain menu item of a certain level may also have multiple menu items below .
- the menu item index contains keywords corresponding to the respective menu items for matching words in the user's voice text;
- the IVR menu in this embodiment is an IVR menu that has been flattened.
- the flattening process means: For the menu items of two adjacent levels in the IVR menu, the menu item of the current level is replaced by the corresponding menu item of the next level.
- a menu item is "Call Charge Inquiry”
- the next menu item includes: “Local Call Charge Inquiry” and “Other Mobile Phone Charge Inquiry”, then the "local call charge inquiry" and "the next level menu item”.
- Other mobile phone bills query "direct instead of "phone bills query” menu items, so that the entire menu tends to be flat, reducing the level of the menu, which helps users quickly reach the required menu items.
- Step S102 providing a menu item matching the user's voice to the user.
- step S102 includes the following process flow: B1: determining whether the number of menu items matching the user is 1, if yes, providing the service logic function of the menu item to the user, otherwise performing step B2;
- the method for quickly entering the IVR menu in the embodiment further includes: Step S103: Extracting different language description information of the same menu item by the user based on the user voice, and establishing a user behavior library to implement self-training of the IVR system, In order to facilitate the subsequent provision of a large language expression range for the user, it is also possible to more accurately identify the menu item that the user wants to reach.
- the second embodiment of the present invention corresponds to the first embodiment. This embodiment introduces a device for quickly entering an IVR menu. As shown in FIG. 2, the following components are included:
- the matching module 100 is configured to perform matching in the IVR menu based on the user voice.
- the matching module 100 includes the following component modules:
- the index establishing module 10 is configured to establish a menu item index based on the text corresponding to the IVR menu.
- the IVR menu is composed of menu items. If there are multiple levels of menu items, there may be multiple menu items under one menu item of a certain level.
- the menu item index contains keywords corresponding to the respective menu items for matching the words in the user's voice text.
- the index establishing module 10 is further configured to: first flatten the IVR menu and then build a menu item index based on the text corresponding to the processed IVR menu.
- the flattening process means: For the menu items of two adjacent levels in the IVR menu, the menu item of the current level is replaced by the corresponding menu item of the next level.
- the total number of levels in the IVR menu can be set as desired, but should be less than the number of levels in the existing IVR menu.
- the flattening of existing IVR menus can also be performed selectively for a specific level or level of menu items as needed.
- the voice text obtaining module 20 is configured to: after the user prompts the menu item function that can be used by the user, record the user voice and convert it into a user voice text, and preprocess the user voice text, and the pre-processed user voice text is The text corresponding to the obtained user's voice.
- the pre-processing includes: de-stop word processing, and use regular expressions to filter out non-GBK-encoded characters and non-text characters.
- the voice text obtaining module 20 is further configured to: segment the pre-processed user voice text, and use the segmented user voice text as the text corresponding to the acquired user voice.
- the word segmentation may be based on word-of-speech or word meaning to decompose the user's speech text to better match the keywords in the menu item index.
- the text matching module 30 is configured to match the text corresponding to the user voice with the index of the IVR menu item to find a menu item that matches the voice of the user.
- the processing module 200 is configured to provide a menu item that matches the user's voice to the user.
- the processing module 200 includes the following component modules:
- the determining module 40 is configured to determine whether the number of menu items matching the user is 1, and if yes, provide the service logic function of the menu item to the user, otherwise invoke the voice guiding module. 50;
- the voice guiding module 50 is configured to perform voice guidance on the user based on the user voice text, prompting the user to input a more accurate voice, and the matching module re-matches.
- the device for quickly entering the IVR menu of the embodiment further includes: 3) a user behavior learning module 300, which extracts different language description information of the user for the same menu item based on the user voice, and establishes a user behavior library to implement the IVR.
- FIG. 3 is a schematic structural diagram of the structure of the IVR system of the embodiment.
- the IVR system includes: a session processing module 101, a voice recognition module 102, a semantic analysis module 103, a user behavior analysis module 104, and a service logic module 105.
- the session processing module 101, the speech recognition module 102, and the semantic analysis module 103 are similar to the functions of the matching module in the second embodiment, and the user behavior analysis module 104 is similar to the function of the user behavior learning module in the second embodiment, business logic.
- Module 105 is similar to the functionality of the processing module in the second embodiment.
- the voice processing module 101 is configured to receive a voice request from the user, control and manage each telephone session, and send the recorded data to the voice recognition module 102; the voice recognition module 102 is configured to recognize the voice input, and convert the recorded data into text.
- the data is sent to the semantic analysis module; the semantic analysis module 103 includes: a menu flattening processing sub-module 106, a user semantic analysis sub-module 107, and a menu routing sub-module 108.
- the menu flattening processing sub-module 106 is configured to flatten the IVR menu to reduce the menu level; the user semantic analysis sub-module 107 is configured to analyze and recognize the natural language. Another
- the user behavior analysis module 104 is configured to establish a user behavior log, extract the user's description of the menu, and store it in the menu information database to implement self-training of the IVR.
- the user service logic module 105 is configured to execute the flow of the IVR business logic and interact with the semantic analysis module 103 to perform various business logic functions.
- the process performed by the semantic analysis module in this embodiment, as shown in FIG. 4, includes: Step S201, initializing, flattening the IVR menu and loading the menu information; Step S202, creating a menu information index, and saving in the computer memory Step S203, preprocessing the speech recognition result, mainly performing de-stopping words on the speech recognition result text and filtering non-GBK characters and non-text characters; and step S204, performing segmentation on the pre-processed speech recognition result text and performing Menu index query and matching; Step S205, determining the number of menu items matched in the menu index query result, if it is 0 or more, then proceeding to step S206, if it is one, proceeding directly to step S207; Step S206, according to the user The previous voice input actively guides the user step by step and obtains the menu item that the user finally needs; Step S207, the IVR system performs a jump
- Step S301 The process starts;
- Step S302 The user dials the IVR system phone number;
- Step S303 The IVR system accesses the user to the system and plays the prompt
- Step S304 The user speaks the menu item that he wants to enter after hearing the prompt tone;
- Step S305 The IVR system receives the user voice and recognizes the voice input.
- Step S306 The IVR system performs semantic analysis on the voice recognition result and obtains the menu analysis result.
- Step S307 The IVR system enters the corresponding menu item according to the menu analysis result, if any The system needs to actively guide the user step by step and obtain the menu item that the user finally needs; Step S308: Jump to the corresponding menu item for the user; Step S309: The process ends.
- the method and device for quickly entering the IVR menu enable the user to quickly reach the menu item in the required IVR menu through natural language interaction, thereby reducing the user's use obstacle to the IVR system, and the specific performance is as follows: Advantages: 1) Directly enter the IVR menu through natural language interaction without using buttons, saving the user's waiting time and improving the IVR experience;
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Epidemiology (AREA)
- General Engineering & Computer Science (AREA)
- Public Health (AREA)
- General Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Primary Health Care (AREA)
- Medical Informatics (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Multimedia (AREA)
- Acoustics & Sound (AREA)
- Human Computer Interaction (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Signal Processing (AREA)
- User Interface Of Digital Computer (AREA)
Abstract
本发明提出了一种快速进入IVR菜单的方法及装置,该方法包括:基于用户语音在IVR菜单中进行匹配,将与用户语音匹配的菜单项提供给用户。该装置包括:匹配模块和处理模块。本发明使用户直接通过自然语言交互以快速到达需要的IVR菜单中的菜单项,节省了用户的等候时间,并提升了IVR的使用体验;优选的,若再加上IVR菜单扁平化处理,减少菜单层级;当出现用户语音输入不太准确时,能够根据用户之前输入的语音,主动对用户进行逐级引导,使用户可以快速进入需要的菜单。
Description
一种快速进入 IVR菜单的方法及装置 技术领域 本发明涉及 IVR (Interactive Voice Response, 互动式语音应答) 系统技术领域, 尤其涉及一种快速进入 IVR菜单的方法及装置。 背景技术
IVR系统是呼叫中心实现人机交互的重要门户, 在传统的 IVR系统中, 用户与系 统交互的方式主要是通过电话的键盘。然而, 随着业务逻辑及菜单复杂度的不断增加, 用户往往需要多次按键或长时间等候菜单提示语音才能进入相应的 IVR菜单项, 从而 使得用户与 IVR系统的交互受到很大的限制, 这也成为当前用户使用体验中最不满意 的地方。 另外, 还有一种情况是, 用户已进入某一级 IVR菜单时不小心按错按键, 可 能导致返回主菜单, 此时再重新自上而下的进行菜单项选择必然浪费用户的时间, 也 极大影响了用户使用 IVR系统的效率。 发明内容 本发明实施例要解决的技术问题是, 提供一种快速进入 IVR菜单的方法及装置, 克服现有技术中通过按键方式进入 IVR菜单项所带来的缺陷。 本发明实施例采用的技术方案是, 所述快速进入 IVR菜单的方法, 包括: 基于用户语音在 IVR菜单中进行匹配, 将与用户语音匹配的菜单项提供给用户。 优选地, 所述基于用户语音在 IVR菜单中进行匹配, 包括: 基于 IVR菜单对应的文本建立菜单项索引; 将用户语音对应的文本与菜单项索引进行匹配, 找出与用户语音匹配的菜单项。 优选地, 所述 IVR菜单为进行过扁平化处理的 IVR菜单; 所述扁平化处理包括: 针对 IVR菜单中两个相邻层级的菜单项, 将本级菜单项由 对应的下一级菜单项代替。 优选地, 所述用户语音对应的文本的获取过程, 包括:
通过语音提示用户可以使用的菜单项功能后, 录制用户语音并将其转换成用户语 音文本, 对所述用户语音文本进行预处理, 所述预处理包括: 去停用词处理、 以及过 滤掉非 GBK (汉字内码扩展规范) 编码的字符和非文字字符。 优选地, 所述用户语音对应的文本的获取过程, 还包括: 对预处理后的用户语音文本进行分词。 优选地, 所述将与用户语音匹配的菜单项提供给用户, 包括:
A1 : 判断与用户匹配的菜单项个数是否为 1, 若是, 则将所述菜单项的业务逻辑 功能提供给用户, 否则执行步骤 A2;
A2: 基于用户语音文本对用户进行语音引导, 提示用户输入更准确的语音, 重新 进行匹配。 本发明实施例还提供一种快速进入 IVR菜单的装置, 包括: 匹配模块, 设置为基于用户语音在 IVR菜单中进行匹配; 处理模块, 设置为将与用户语音匹配的菜单项提供给用户。 优选地, 所述匹配模块, 包括: 索引建立模块, 设置为基于 IVR菜单对应的文本建立菜单项索引; 文本匹配模块, 设置为将用户语音对应的文本与菜单项索引进行匹配, 找出与用 户语音匹配的菜单项。 优选地, 所述匹配模块, 包括: 菜单处理模块, 设置为对所述 IVR菜单先进行扁平化处理后再提供给匹配模块; 所述扁平化处理包括: 针对 IVR菜单中两个相邻层级的菜单项, 将本级菜单项由 对应的下一级菜单项代替。 本发明还提供一种 IVR系统, 包括: 录音设备, 设置为录制用户语音并将其转换成用户语音文本;
处理器, 设置为基于 IVR菜单对应的文本建立菜单项索引, 将用户语音文本与菜 单项索引进行匹配, 找出与用户语音匹配的菜单项, 将所述菜单项的业务逻辑功能提 供给用户。 采用上述技术方案, 本发明实施例至少具有下列优点: 本发明实施例所述快速进入 IVR菜单的方法及装置, 使用户直接通过自然语言交 互以快速到达需要的 IVR菜单中的菜单项, 节省了用户的等候时间, 并提升了 IVR的 使用体验; 优选的, 若再加上 IVR菜单扁平化处理, 减少菜单层级; 当出现用户语音 输入不太准确时, 能够根据用户之前输入的语音, 主动对用户进行逐级引导, 使用户 可以快速进入需要的菜单。 附图说明 图 1为本发明第一实施例的快速进入 IVR菜单的方法流程图; 图 2为本发明第二实施例的快速进入 IVR菜单的装置组成结构示意图; 图 3为本发明第三实施例的 IVR系统组成结构示意图; 图 4为本发明第三实施例的语义分析模块执行的流程图; 图 5为本发明第三实施例的快速进入 IVR菜单的流程图。 具体实施方式 为更进一步阐述本发明为达成预定目的所采取的技术手段及功效, 以下结合附图 及较佳实施例, 对本发明进行详细说明如后。 本发明第一实施例, 一种快速进入 IVR菜单的方法, 如图 1所示, 包括以下具体 步骤: 步骤 S101 , 基于用户语音在 IVR菜单中进行匹配。 具体的, 步骤 S101包括以下处理流程: A1 : 获取用户语音对应的文本; 进一步的, 通过语音提示用户可以使用的菜单项功能后, 录制用户语音并将其转 换成用户语音文本, 对该用户语音文本进行预处理, 该预处理包括: 去停用词处理、
以及利用正则表达式过滤掉非 GBK编码的字符和非文字字符, 典型的非文字字符, 例如: 标点符号、 以及& #等字符, 去停用词可以基于停用词词典进行。 预处理后的 用户语音文本即为获取到的用户语音对应的文本。 优选的, 对预处理后的用户语音文本还要进行分词, 将经过分词后的用户语音文 本作为获取到的用户语音对应的文本。 分词可以是根据词性或者词义等对用户语音文 本进行词语分解, 以便更好的与菜单项索引中的关键词匹配上。
A2: 基于 IVR菜单对应的文本建立菜单项索引; 优选地, IVR菜单是由各菜单项组成的, 如果有多级菜单项, 那么某级的某个菜 单项下面可能还对应有多个菜单项。 菜单项索引包含与各菜单项相对应的关键词, 用 于与用户语音文本中的词语相匹配;
A3: 将用户语音对应的文本与 IVR菜单项索引进行匹配, 找出与用户语音匹配的 菜单项。 优选的, 本实施例中的 IVR菜单为进行过扁平化处理的 IVR菜单。 该扁平化处理是指: 针对 IVR菜单中两个相邻层级的菜单项, 将本级菜单项由对 应的下一级菜单项代替。 比如: 某一菜单项是"话费查询", 其下一级菜单项包括: "本 机话费查询"和 "其他手机话费查询", 则将下一级菜单项的 "本机话费查询"和 "其 他手机话费查询"直接代替 "话费查询"菜单项, 使整个菜单趋于扁平化, 减少菜单 的层级, 有利于用户快速到达需要的菜单项。
IVR菜单中菜单的总级数可以根据需要设置,但应当少于现有 IVR菜单中的级数。 对于现有 IVR菜单的扁平化处理也可以根据需要, 有选择性的针对特定的某级或者某 几级菜单项来进行。 当经过扁平化处理使得 IVR菜单的总级数缩减为一级时, 执行本 发明实施例的技术方案, 能使用户最快达到所需的菜单项, 实现本发明的最佳效果。 步骤 S102, 将与用户语音匹配的菜单项提供给用户。 具体的, 步骤 S 102包括以下处理流程: B1 : 判断与用户匹配的菜单项个数是否为 1, 若是, 则将所述菜单项的业务逻辑 功能提供给用户, 否则执行步骤 B2;
B2: 基于用户语音文本对用户进行语音引导, 提示用户输入更准确的语音, 重新 进行匹配。
优选的, 本实施例的所述快速进入 IVR菜单的方法, 还包括: 步骤 S103 , 基于用户语音提取用户对于相同菜单项的不同语言描述信息, 建立用 户行为库, 以实现 IVR系统的自我训练, 便于后续为用户提供较大的语言表述范围的 情况下, 还能较准确的识别出用户想要达到的菜单项。 本发明第二实施例, 与第一实施例对应, 本实施例介绍一种快速进入 IVR菜单的 装置, 如图 2所示, 包括以下组成部分:
1 ) 匹配模块 100, 设置为基于用户语音在 IVR菜单中进行匹配; 具体的, 匹配模块 100包括以下组成模块: 索引建立模块 10, 设置为基于 IVR菜单对应的文本建立菜单项索引。 IVR菜单是 由各菜单项组成的, 如果有多级菜单项, 那么某级的某个菜单项下面可能还对应有多 个菜单项。 菜单项索引包含与各菜单项相对应的关键词, 用于与用户语音文本中的词 语相匹配。 优选的, 索引建立模块 10, 还设置为: 对所述 IVR菜单先进行扁平化处理后再基 于处理后的 IVR菜单对应的文本建立菜单项索引。 该扁平化处理是指: 针对 IVR菜单中两个相邻层级的菜单项, 将本级菜单项由对 应的下一级菜单项代替。 IVR菜单中菜单的总级数可以根据需要设置, 但应当少于现 有 IVR菜单中的级数。对于现有 IVR菜单的扁平化处理也可以根据需要, 有选择性的 针对特定的某级或者某几级菜单项来进行。 语音文本获取模块 20, 设置为通过语音提示用户可以使用的菜单项功能后, 录制 用户语音并将其转换成用户语音文本, 对所述用户语音文本进行预处理, 预处理后的 用户语音文本即为获取到的用户语音对应的文本。 该预处理包括: 去停用词处理、 以 及利用正则表达式过滤掉非 GBK编码的字符和非文字字符。 优选的, 语音文本获取 模块 20, 还设置为: 对预处理后的用户语音文本进行分词, 将经过分词后的用户语音 文本作为获取到的用户语音对应的文本。 分词可以是根据词性或者词义等对用户语音 文本进行词语分解, 以便更好的与菜单项索引中的关键词匹配上。 文本匹配模块 30, 设置为将用户语音对应的文本与 IVR菜单项索引进行匹配, 找 出与用户语音匹配的菜单项。
2) 处理模块 200, 设置为将与用户语音匹配的菜单项提供给用户。
具体的, 处理模块 200包括以下组成模块: 判断模块 40, 设置为判断与用户匹配的菜单项个数是否为 1, 若是, 则将该菜单 项的业务逻辑功能提供给用户, 否则调用语音引导模块 50; 语音引导模块 50, 设置为基于用户语音文本对用户进行语音引导, 提示用户输入 更准确的语音, 供匹配模块重新进行匹配。 比如: 用户语音是"我想查询", 由于 IVR 菜单提供给用户查询的项目有很多种, 此时基于用户语音文本对用户进行语音引导时 可以播报 "查询话费"、 "查询套餐流量"、 "查询余额"供用户从中选择并通过自然语 言反馈用户的选择, 重新调用匹配模块 100进行匹配。 优选的, 本实施例的所述快速进入 IVR菜单的装置, 还包括: 3 )用户行为学习模块 300, 基于用户语音提取用户对于相同菜单项的不同语言描 述信息, 建立用户行为库, 以实现 IVR系统的自我训练, 可用于指导匹配模块 100, 便于后续为用户提供较大的语言可变表述范围的情况下, 还能较准确的识别出用户想 要达到的菜单项。 本发明第三实施例, 本实施例是在上述实施例的基础上, 结合图 3~5介绍一个本 发明的应用实例。 图 3为本实施例的 IVR系统组成结构示意图。 如图 3所示, 该 IVR系统包括: 话 路处理模块 101、 语音识别模块 102、 语义分析模块 103、 用户行为分析模块 104及业 务逻辑模块 105。 话路处理模块 101、 语音识别模块 102和语义分析模块 103类似于第二实施例中 的匹配模块的功能, 用户行为分析模块 104类似于第二实施例中的用户行为学习模块 的功能, 业务逻辑模块 105类似于第二实施例中的处理模块的功能。 话路处理模块 101设置为接收用户的语音请求, 控制和管理各电话话路, 并将录 音数据送入语音识别模块 102; 语音识别模块 102设置为对语音输入进行识别, 即将录音数据转换成文本数据, 并将识别结果送入语义分析模块; 语义分析模块 103包括: 菜单扁平化处理子模块 106、 用户语义分析子模块 107 和菜单路由子模块 108。 其中, 菜单扁平化处理子模块 106设置为对 IVR菜单进行扁 平化处理, 减少菜单层级; 用户语义分析子模块 107设置为对自然语言进行解析及识
另 |J, 并获取最终用于查询菜单项的信息; 菜单路由子模块 108设置为将用户接入到对 应的业务菜单项中。 用户行为分析模块 104设置为建立用户行为日志,提取用户关于菜单的诉求描述, 存入菜单信息库, 以实现 IVR的自我训练。 用户 业务逻辑模块 105设置为执行 IVR业务逻辑的流程, 并与语义分析模块 103进行 交互完成各个业务逻辑功能。 本实施例中语义分析模块执行的流程, 如图 4所示, 包括: 步骤 S201 , 初始化, 对 IVR菜单进行扁平化处理并加载菜单信息; 步骤 S202, 创建菜单信息索引, 并保存在计算机内存中; 步骤 S203 , 对语音识别结果进行预处理, 主要是对语音识别结果文本进行去停用 词及过滤非 GBK字符和非文字字符; 步骤 S204, 对预处理后的语音识别结果文本进行分词并进行菜单索引查询及匹 配; 步骤 S205, 判断菜单索引查询结果中匹配到的菜单项个数, 若为 0个或多个则进 入步骤 S206, 若为 1个则直接进入步骤 S207; 步骤 S206,根据用户之前的语音输入主动对用户进行逐级引导并得到用户最终需 要的菜单项; 步骤 S207, IVR系统根据匹配到的菜单项进行跳转; 步骤 S208, 分析结束。 本实施例的快速进入 IVR菜单的流程如图 5所示, 包括以下步骤: 步骤 S301 : 流程开始; 步骤 S302: 用户拨打 IVR系统电话号码; 步骤 S303 : IVR系统接入用户到系统中并播放提示音; 步骤 S304: 用户听到提示音后说出想要进入的菜单项;
步骤 S305: IVR系统接收用户语音并对语音输入进行识别; 步骤 S306: IVR系统对语音识别结果进行语义分析并得到菜单分析结果; 步骤 S307: IVR系统根据菜单分析结果进入对应的菜单项, 如有需要系统会主动 对用户进行逐级引导并得到用户最终需要的菜单项; 步骤 S308: 为用户跳转到对应的菜单项; 步骤 S309: 流程结束。 本发明所述快速进入 IVR菜单的方法及装置, 使用户直接通过自然语言交互以快 速到达需要的 IVR菜单中的菜单项, 减少了用户对 IVR系统的使用障碍, 具体表现为 以下四个方面的优势: 1 )不使用按键而直接通过自然语言交互进入 IVR菜单, 节省了用户的等候时间, 并提升了 IVR的使用体验;
2) IVR菜单扁平化, 菜单层级减少, 并能够主动对用户进行逐级引导, 使用户可 以快速进入需要的菜单;
3 ) 能够根据用户行为实现 IVR的自我训练, 进而提高 IVR的智能化。 4) 避免用户对按键的误操作导致交互时间延长的情况发生。 通过具体实施方式的说明, 应当可对本发明为达成预定目的所采取的技术手段及 功效得以更加深入且具体的了解, 然而所附图示仅是提供参考与说明之用, 并非用来 对本发明加以限制。 工业实用性 本发明提供的上述技术方案, 可以应用于快速进入互动式语音应答 IVR菜单过程 中, 采用基于用户语音在 IVR菜单中进行匹配, 将与用户语音匹配的菜单项提供给用 户的技术方案,使用户直接通过自然语言交互以快速到达需要的 IVR菜单中的菜单项, 节省了用户的等候时间, 并提升了 IVR的使用体验; 优选的, 若再加上 IVR菜单扁平 化处理, 减少菜单层级; 当出现用户语音输入不太准确时, 能够根据用户之前输入的 语音, 主动对用户进行逐级引导, 使用户可以快速进入需要的菜单。
Claims
权 利 要 求 书 、 一种快速进入互动式语音应答 IVR菜单的方法, 包括: 基于用户语音在 IVR菜单中进行匹配,将与用户语音匹配的菜单项提供给 用户。 、 根据权利要求 1所述的快速进入 IVR菜单的方法, 其中, 所述基于用户语音在 IVR菜单中进行匹配, 包括: 基于 IVR菜单对应的文本建立菜单项索引; 将用户语音对应的文本与菜单项索引进行匹配, 找出与用户语音匹配的菜 单项。 、 根据权利要求 1或 2所述的快速进入 IVR菜单的方法,其中,所述 IVR菜单为 进行过扁平化处理的 IVR菜单; 所述扁平化处理包括: 针对 IVR菜单中两个相邻层级的菜单项, 将本级菜 单项由对应的下一级菜单项代替。 、 根据权利要求 2所述的快速进入 IVR菜单的方法, 其中, 所述用户语音对应的 文本的获取过程, 包括: 通过语音提示用户可以使用的菜单项功能后, 录制用户语音并将其转换成 用户语音文本, 对所述用户语音文本进行预处理, 所述预处理包括: 去停用词 处理、 以及过滤掉非汉字内码扩展规范 GBK编码的字符和非文字字符。 、 根据权利要求 4所述的快速进入 IVR菜单的方法, 其中, 所述用户语音对应的 文本的获取过程, 还包括:
对预处理后的用户语音文本进行分词。 、 根据权利要求 2所述的快速进入 IVR菜单的方法, 其中, 所述将与用户语音匹 配的菜单项提供给用户, 包括:
A1 : 判断与用户语音匹配的菜单项个数是否为 1, 若是, 则将所述菜单项 的业务逻辑功能提供给用户, 否则执行步骤 A2;
A2:基于用户语音文本对用户进行语音引导,提示用户输入更准确的语音, 重新进行匹配。
、 一种快速进入 IVR菜单的装置, 包括: 匹配模块, 设置为基于用户语音在 IVR菜单中进行匹配; 处理模块, 设置为将与用户语音匹配的菜单项提供给用户。 、 根据权利要求 7所述的快速进入 IVR菜单的装置,其中,所述匹配模块,包括: 索引建立模块, 设置为基于 IVR菜单对应的文本建立菜单项索引; 文本匹配模块, 设置为将用户语音对应的文本与菜单项索引进行匹配, 找 出与用户语音匹配的菜单项。 、 根据权利要求 7或 8所述的快速进入 IVR菜单的装置, 其中, 所述匹配模块, 包括: 菜单处理模块, 设置为对所述 IVR菜单先进行扁平化处理后再提供给匹配 模块;
所述扁平化处理包括: 针对 IVR菜单中两个相邻层级的菜单项, 将本级菜 单项由对应的下一级菜单项代替。 0、 一种 IVR系统, 包括:
录音设备, 设置为录制用户语音并将其转换成用户语音文本; 处理器, 设置为基于 IVR菜单对应的文本建立菜单项索引, 将用户语音文 本与菜单项索引进行匹配, 找出与用户语音匹配的菜单项, 将所述菜单项的业 务逻辑功能提供给用户。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP14894447.3A EP3157236A4 (en) | 2014-06-11 | 2014-08-21 | Method and device for quickly accessing ivr menu |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410256760.9A CN105323392A (zh) | 2014-06-11 | 2014-06-11 | 一种快速进入ivr菜单的方法及装置 |
CN201410256760.9 | 2014-06-11 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2015188454A1 true WO2015188454A1 (zh) | 2015-12-17 |
Family
ID=54832778
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2014/084951 WO2015188454A1 (zh) | 2014-06-11 | 2014-08-21 | 一种快速进入ivr菜单的方法及装置 |
Country Status (3)
Country | Link |
---|---|
EP (1) | EP3157236A4 (zh) |
CN (1) | CN105323392A (zh) |
WO (1) | WO2015188454A1 (zh) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111290677A (zh) * | 2018-12-07 | 2020-06-16 | 中电长城(长沙)信息技术有限公司 | 一种自助设备导航方法及其导航系统 |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109935342A (zh) * | 2017-12-15 | 2019-06-25 | 北京医联蓝卡在线科技有限公司 | 一种家庭医生助理帮助系统及方法 |
CN109660678A (zh) * | 2018-12-07 | 2019-04-19 | 深圳前海微众银行股份有限公司 | 电核系统实现方法、系统及可读存储介质 |
CN109979451A (zh) * | 2019-03-20 | 2019-07-05 | 百度在线网络技术(北京)有限公司 | 用于输出信息的方法和装置 |
CN111031185A (zh) * | 2019-12-19 | 2020-04-17 | 易谷网络科技股份有限公司 | 一种基于人工智能导航的坐席分配方法及相关装置 |
CN111081245A (zh) * | 2019-12-24 | 2020-04-28 | 杭州纪元通信设备有限公司 | 一种基于语音识别的呼叫中心的菜单系统 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101609673A (zh) * | 2009-07-09 | 2009-12-23 | 交通银行股份有限公司 | 一种基于电话银行的用户语音处理方法及服务器 |
CN102300007A (zh) * | 2010-06-23 | 2011-12-28 | 上海博路信息技术有限公司 | 一种基于语音识别的呼叫中心的扁平化菜单系统 |
CN103002156A (zh) * | 2011-12-16 | 2013-03-27 | 微软公司 | 用于动态语音菜单的技术 |
CN103188407A (zh) * | 2011-12-31 | 2013-07-03 | 中国移动通信集团广东有限公司 | 交互式语音应答ivr的处理方法、终端、拨测服务器和系统 |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7797159B2 (en) * | 2002-09-16 | 2010-09-14 | Movius Interactive Corporation | Integrated voice navigation system and method |
US8457973B2 (en) * | 2006-03-04 | 2013-06-04 | AT&T Intellectual Propert II, L.P. | Menu hierarchy skipping dialog for directed dialog speech recognition |
US8185539B1 (en) * | 2008-08-12 | 2012-05-22 | Foneweb, Inc. | Web site or directory search using speech recognition of letters |
US8321277B2 (en) * | 2008-06-18 | 2012-11-27 | Nuance Communications, Inc. | Method and system for voice ordering utilizing product information |
-
2014
- 2014-06-11 CN CN201410256760.9A patent/CN105323392A/zh not_active Withdrawn
- 2014-08-21 WO PCT/CN2014/084951 patent/WO2015188454A1/zh active Application Filing
- 2014-08-21 EP EP14894447.3A patent/EP3157236A4/en not_active Withdrawn
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101609673A (zh) * | 2009-07-09 | 2009-12-23 | 交通银行股份有限公司 | 一种基于电话银行的用户语音处理方法及服务器 |
CN102300007A (zh) * | 2010-06-23 | 2011-12-28 | 上海博路信息技术有限公司 | 一种基于语音识别的呼叫中心的扁平化菜单系统 |
CN103002156A (zh) * | 2011-12-16 | 2013-03-27 | 微软公司 | 用于动态语音菜单的技术 |
CN103188407A (zh) * | 2011-12-31 | 2013-07-03 | 中国移动通信集团广东有限公司 | 交互式语音应答ivr的处理方法、终端、拨测服务器和系统 |
Non-Patent Citations (1)
Title |
---|
See also references of EP3157236A4 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111290677A (zh) * | 2018-12-07 | 2020-06-16 | 中电长城(长沙)信息技术有限公司 | 一种自助设备导航方法及其导航系统 |
CN111290677B (zh) * | 2018-12-07 | 2023-09-19 | 中电长城(长沙)信息技术有限公司 | 一种自助设备导航方法及其导航系统 |
Also Published As
Publication number | Publication date |
---|---|
EP3157236A1 (en) | 2017-04-19 |
CN105323392A (zh) | 2016-02-10 |
EP3157236A4 (en) | 2017-06-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105931644B (zh) | 一种语音识别方法及移动终端 | |
KR101768509B1 (ko) | 온라인 음성 번역 방법 및 장치 | |
CN110444198B (zh) | 检索方法、装置、计算机设备和存储介质 | |
US9350862B2 (en) | System and method for processing speech | |
JP6182272B2 (ja) | 自然表現の処理方法、処理及び応答方法、装置、及びシステム | |
CN106373575B (zh) | 一种用户声纹模型构建方法、装置及系统 | |
US7450698B2 (en) | System and method of utilizing a hybrid semantic model for speech recognition | |
JP4089148B2 (ja) | 通訳サービス方法および通訳サービス装置 | |
WO2015188454A1 (zh) | 一种快速进入ivr菜单的方法及装置 | |
KR102445519B1 (ko) | 대화형 지능 서비스 제공 챗봇 제작 시스템 및 방법 | |
CN110164435A (zh) | 语音识别方法、装置、设备及计算机可读存储介质 | |
CN109086276B (zh) | 数据翻译方法、装置、终端及存储介质 | |
KR20170033722A (ko) | 사용자의 발화 처리 장치 및 방법과, 음성 대화 관리 장치 | |
CN103853703A (zh) | 一种信息处理方法及电子设备 | |
CN109979450B (zh) | 信息处理方法、装置及电子设备 | |
CN111949255A (zh) | 基于语音的脚本编写方法、装置、设备及存储介质 | |
CN106713111B (zh) | 一种添加好友的处理方法、终端及服务器 | |
CN111142834A (zh) | 一种业务处理方法及系统 | |
CN106371905B (zh) | 应用程序操作方法、装置和服务器 | |
KR101207435B1 (ko) | 대화형 음성인식 서버, 대화형 음성인식 클라이언트 및 대화형 음성인식 방법 | |
CN109388695B (zh) | 用户意图识别方法、设备及计算机可读存储介质 | |
CN113936655A (zh) | 语音播报处理方法、装置、计算机设备和存储介质 | |
JP7211384B2 (ja) | 音声認識装置、個人識別方法および個人識別プログラム | |
TWI752474B (zh) | 無障礙智能語音系統及其控制方法 | |
TWI242729B (en) | Speech database establishment and recognition method and system thereof |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 14894447 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
REEP | Request for entry into the european phase |
Ref document number: 2014894447 Country of ref document: EP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2014894447 Country of ref document: EP |