披露设计模式Disclosure design patterns

你已经确定了合成语音体验的正确披露级别,现在正是探索潜在设计模式的好时机。Now that you've determined the right level of disclosure for your synthetic voice experience, it's a good time to explore potential design patterns.

概述Overview

下面是可应用于合成语音体验的披露设计模式范围。There's a spectrum of disclosure design patterns you can apply to your synthetic voice experience. 如果披露评估结果为“高披露”,建议显式披露,即直接传达合成语音的来源。If the outcome of your disclosure assessment was 'High Disclosure', we recommend explicit disclosure, which means communicating the origins of the synthetic voice outright. 隐式披露包括提示和交互模式,无论所需披露级别是高是低,这些提示和交互模式都会使语音体验受益。Implicit disclosure includes cues and interaction patterns that benefit voice experiences whether or not required disclosure levels are high or low. 披露模式范围Spectrum of disclosure patterns

显式披露模式Explicit disclosure patterns 隐式披露模式Implicit disclosure patterns
透明介绍Transparent Introduction
言语透明介绍Verbal Transparent Introduction
显式署名Explicit Byline
自定义和校准Customization and Calibration
向家长披露Parental Disclosure
提供详细了解语音发出方式的机会Providing opportunities to learn more about how the voice was made
功能披露Capability Disclosure
隐式提示和反馈Implicit Cues and Feedback
对话透明度Conversational Transparency

使用以下图表,直接参考适用于合成语音的模式。Use the following chart to refer directly to the patterns that apply to your synthetic voice. 此图表中的其他一些条件可能也适用于你的方案:Some of the other conditions in this chart may also apply to your scenario:

如果你的合成语音体验…If your synthetic voice experience… 建议Recommendations 设计模式Design patterns
需要高披露Requires High Disclosure 预先使用至少一种显式模式和隐式提示,以帮助用户建立关联。Use at least one explicit pattern and implicit cues up front to helps users build associations. 显式披露Explicit Disclosure
隐式披露Implicit Disclosure
需要低披露Requires Low Disclosure 披露可能很少或不必要,但可从某些隐式模式中受益。Disclosure may be minimal or unnecessary, but could benefit from some implicit patterns. 功能披露Capability Disclosure
对话透明度Conversational Transparency
具有高参与度Has a high level of engagement 为长期使用而构建,提供多个入口点,以在用户旅程中进行披露。Build for the long term and offer multiple entry points to disclosure along the user journey. 强烈建议设计载入体验。It is highly recommended to have an onboarding experience. 透明介绍Transparent Introduction
自定义和校准Customization and Calibration
功能披露Capability Disclosure
包含儿童作为主要目标受众Includes children as the primary intended audience 将家长作为主要披露受众,并确保他们能够有效地将披露传达给孩子。Target parents as the primary disclosure audience and ensure that they can effectively communicate disclosure to children. 向家长披露Parental Disclosure
言语透明介绍Verbal Transparent Introduction
隐式披露Implicit Disclosure
对话透明度Conversational Transparency
包含盲人用户或低视力人群作为主要目标受众Includes blind users or people with low vision as the primary intended audience 包含所有用户,并确保任何形式的视觉披露都具有关联的替换文字或声音效果。Be inclusive of all users and ensure that any form of visual disclosure has associated alternative text or sound effects. 遵循对比度和显示大小的辅助功能标准。Adhere to accessibility standards for contrast ratio and display size. 使用听觉提示传达披露。Use auditory cues to communicate disclosure. 言语透明介绍Verbal Transparent Introduction
听觉提示Auditory Cues
触觉提示Haptic Cues
对话透明度Conversational Transparency
辅助功能标准Accessibility Standards
无屏幕、无设备,或使用语音作为主要或唯一交互模式Is screen-less, device-less or uses voice as the primary or only mode of interaction 使用听觉提示传达披露。Use auditory cues to communicate disclosure. 言语透明介绍Verbal Transparent Introduction
听觉提示Auditory Cues
可能包含多个用户/听众(例如多个家庭中的私人助理)Potentially includes multiple users/listeners (e.g., personal assistant in multiple household) 请注意各种用户上下文和理解级别,并在用户旅程中提供多个披露机会。Be mindful of various user contexts and levels of understanding and offer multiple opportunities for disclosure in the user journey. 透明介绍(返回用户)Transparent Introduction (Return User)
提供详细了解语音发出方式的机会Providing opportunities to learn more about how the voice was made
对话透明度Conversational Transparency

显式披露Explicit disclosure

如果合成语音体验需要高披露,则最好使用以下显式模式中的至少一种来清楚地声明合成性质。If your synthetic voice experience requires high disclosure, it's best to use at least one of the following explicit patterns to clearly state the synthetic nature.

透明介绍Transparent Introduction

在开始语音体验之前,请介绍数字助理,完全透明地公开其语音来源和功能。Before the voice experience begins, introduce the digital assistant by being fully transparent about the origins of its voice and its capabilities. 加入新用户或向返回用户介绍新功能时是使用此模式的最佳时机。The optimal moment to use this pattern is when onboarding a new user or when introducing new features to a returning user. 在介绍过程中实现隐式提示,可帮助用户形成有关数字代理合成性质的心理模型。Implementing implicit cues during an introduction helps users form a mental model about the synthetic nature of the digital agent.

首次用户体验First-time user experience

首次运行体验过程中的透明介绍Transparent introduction during first run experience
合成语音在加入新用户时引入。The synthetic voice is introduced while onboarding a new user.

建议Recommendations

  • 说明语音是人工合成的(例如“数字”)Describe that the voice is artificial (e.g. "digital")
  • 说明代理能够执行的操作Describe what the agent is capable of doing
  • 显式说明语音的来源Explicitly state the voice's origins
  • 提供可详细了解合成语音的入口点Offer an entry point to learn more about the synthetic voice

返回用户体验Returning user experience

如果用户跳过载入体验,继续提供透明介绍体验的入口点,直到用户首次触发语音。If a user skips the onboarding experience, continue to offer entry points to the Transparent Introduction experience until the user triggers the voice for the first time.

返回用户体验过程中的透明介绍Transparent introduction during return user experience
提供合成语音体验的一致入口点。当用户在用户旅程中的任何时间点首次触发语音时,允许其返回载入体验。Provide a consistent entry point to the synthetic voice experience. Allow the user to return to the onboarding experience when they trigger the voice for the first time at any point in the user journey.

言语透明介绍Verbal Transparent Introduction

通过语音提示陈述数字助理的语音来源,足以显式实现披露。A spoken prompt stating the origins of the digital assistant's voice is explicit enough on its own to achieve disclosure. 此模式最适用于高披露方案,在此类方案中,语音是唯一可用的交互模式。This pattern is best for high disclosure scenarios where voice is the only mode of interaction available.

口述透明介绍 Verbally spoken transparent introduction
在用户体验中的某些时刻需要介绍或说明某人的音源时,请使用透明介绍。Use a transparent introduction when there are moments in the user experience where you might already introduce or attribute a person's voice.

第一人称口述透明介绍Verbally spoken transparent introduction in first person
为提高透明度,配音演员可以以第一人称披露合成语音的来源。For additional transparency, the voice actor can disclose the origins of the synthetic voice in the first person.

显式署名Explicit Byline

如果用户将与音频播放器或交互式组件进行交互以触发语音,请使用此模式。Use this pattern if the user will be interacting with an audio player or interactive component to trigger the voice.

新闻媒体方案中的显式署名Explicit byline in a news media scenario
显式署名是指说明语音来源An explicit byline is the attribution of where voice came from.

建议Recommendations

  • 提供详细了解合成语音的入口点Offer entry point to learn more about the synthesized voice

自定义和校准Customization and Calibration

为用户提供对数字助理对其响应方式(即语音的听感)的控制。Provide users control over how the digital assistant responds to them (i.e., how the voice sounds). 如果用户以其自己的方式与系统进行交互,且在与系统交互时心中有特定目标,则根据定义,他们已经认识到系统不是真人。When a user interacts with a system on their own terms and with specific goals in mind, then by definition, they have already understood that it's not a real person.

用户控件User Control

提供对合成语音体验具有意义深刻且显著影响的选择。Offer choices that have a meaningful and noticeable impact on the synthetic voice experience.

用户首选项User preferences
用户首选项允许用户自定义及改进其体验。User preferences allow users to customize and improve their experience.

建议Recommendations

  • 允许用户自定义语音(如选择语言和语音类型)Allow users to customize the voice (e.g., select language and voice type)
  • 向用户提供一种教授系统响应其独特语音的方法(如语音校准、自定义命令)Provide users a way to teach the system to respond to his/her unique voice (e.g., voice calibration, custom commands)
  • 优化用户生成的交互或上下文交互(如提醒)Optimize for user-generated or contextual interactions (e.g., reminders)

角色自定义Persona Customization

提供自定义数字助理语音的方法。Offer ways to customize the digital assistant's voice. 如果语音基于公众人物或知名度较高的人,请考虑在用户预览语音时同时使用视觉和语音介绍。If the voice is based on a celebrity or a widely recognizable person, consider using both visual and spoken introductions when users preview the voice.

语音自定义Voice customization
提供从一组语音中进行选择的功能,有助于表达人工合成这一性质。Offering the ability to select from a set of voices helps convey the artificial nature.

建议Recommendations

  • 允许用户预览每个语音的声音Allow users to preview the sound of each voice
  • 为每个语音使用真实的介绍Use an authentic introduction for each voice
  • 提供详细了解合成语音的入口点Offer entry points to learn more about the synthesized voice

向家长披露Parental Disclosure

除遵守 COPPA 法规外,如果你的主要目标受众是少年儿童,且公开水平较高,则请向家长提供披露。In addition to complying with COPPA regulations, provide disclosure to parents if your primary intended audience is young children and your exposure level is high. 对于敏感用途,请考虑在成人认可使用合成语音前管控这一体验。For sensitive uses, consider gating the experience until an adult has acknowledged the use of the synthetic voice. 鼓励家长向其孩子传达消息。Encourage parents to communicate the message to their children.

向家长披露Disclosure for parents
针对家长进行了优化的透明介绍,确保在孩子与语音交互之前,成人已经了解该语音的合成性质。A transparent introduction optimized for parents ensures that an adult was made aware of the synthetic nature of the voice before a child interacts with it.

建议Recommendations

  • 以家长为主要受众进行披露Target parents as the primary audience for disclosure
  • 鼓励家长向其孩子传达披露Encourage parents to communicate disclosure to their children
  • 提供详细了解合成语音的入口点Offer entry points to learn more about the synthesized voice
  • 通过向家长询问一个简单的“保护”问题来证明他们已经阅读了披露内容,从而管控体验Gate the experience by asking parents a simple "safeguard" question to show they have read the disclosure

提供详细了解语音发出方式的机会Providing opportunities to learn more about how the voice was made

提供上下文敏感入口点,指向可提供有关合成语音技术详细信息的页面、弹出项或外部站点。Offer context-sensitive entry points to a page, pop-up, or external site that provides more information about the synthetic voice technology. 例如,在载入过程中或在对话过程中用户表示需要更多信息时,可以显示链接供其了解详细信息。For example, you could surface a link to learn more during onboarding or when the user prompts for more information during conversation.

了解详细信息的入口点Entry point to learn more
提供详细了解合成语音机会的输入点的示例。Example of an entry point to offer the opportunity to learn more about the synthesized voice.

用户请求有关合成语音的详细信息后,主要目标是向他们讲解合成语音的来源,并透明地介绍该技术。Once a user requests more information about the synthetic voice, the primary goal is to educate them about the origins of the synthetic voice and to be transparent about the technology.

向用户提供有关合成语音的详细信息Provide users more information about synthetic voice
可在外部站点帮助站点中提供详细信息。More information can be offered in an external site help site.

建议Recommendations

  • 简化复杂概念,避免使用法律术语和技术术语Simplify complex concepts and avoid using legalese and technical jargon
  • 请勿将此内容埋没在隐私声明和使用条款中Don't bury this content in privacy and terms of use statements
  • 保持内容简洁,在可用时使用图像Keep content concise and use imagery when available

隐式披露Implicit disclosure

一致性是在整个用户旅程中隐式实现披露的关键。Consistency is the key to achieving disclosure implicitly throughout the user journey. 跨设备和交互模式一致地使用视觉和听觉提示,有助于在隐式模式和显式披露之间建立关联。Consistent use of visual and auditory cues across devices and modes of interaction can help build associations between implicit patterns and explicit disclosure.

隐式提示的一致性

隐式提示和反馈Implicit Cues & Feedback

可通过不同的方式表达拟人形象,从代理的实际视觉表示形式到语音、声音、光模式、弹跳形状甚至设备振动。Anthropomorphism can manifest in different ways, from the actual visual representation of the agent, to the voice, sounds, patterns of light, bouncing shapes, or even the vibration of a device. 定义角色时,请利用隐式提示和反馈模式,而不要试图打造很像真人的头像。When defining your persona, leverage implicit cues and feedback patterns rather than aim for a very human-like avatar. 这是最大程度减少对更显式披露需求的一种方法。This is one way to minimize the need for more explicit disclosure.

视觉提示和反馈Visual cues and feedback
这些提示可帮助使代理拟人化而又不过于像真人。长期一致地使用,它们本身也可以成为有效的披露机制。These cues help anthropomorphize the agent without being too human-like. They can also become effective disclosure mechanisms on their own when used consistently over time.

结合以下类型的提示时,请考虑体验的不同交互模式:Consider the different modes of interactions of your experience when incorporating the following types of cues:

视觉提示Visual Cues 听觉提示Auditory Cues 触觉提示Haptic Cues
AvatarAvatar
响应式实时提示(如动画)Responsive real-time cues (e.g., animations)
非屏幕提示(如设备上的光和图案)Non-screen cues (e.g., lights and patterns on a device)
Sonicon(如简短的独特声音、一串音符)Sonicon (e.g., a brief distinctive sound, series of musical notes) 振动Vibration

功能披露Capability Disclosure

可通过对数字助理能够执行的操作设置准确期望来隐式实现披露。Disclosure can be achieved implicitly by setting accurate expectations for what the digital assistant is capable of. 提供示例命令,以便用户可以了解如何与数字助理交互,并提供上下文帮助,以在体验的早期阶段了解有关合成语音的详细信息。Provide sample commands so that users can learn how to interact with the digital assistant and offer contextual help to learn more about the synthetic voice during the early stages of the experience.

视觉提示和反馈

对话透明度Conversational Transparency

当会话进入意外路径时,请考虑创建可帮助重置预期、增强透明度,以及将用户导向正确路径的默认响应。When conversations fall in unexpected paths, consider crafting default responses that can help reset expectations, reinforce transparency, and steer users towards successful paths. 也有机会在对话中使用显式披露。There are opportunities to use explicit disclosure in conversation as well.

处理意外路径


指向代理的任务外或“私人”问题,是提醒用户代理的合成性质并引导用户适当与之交流或将用户重定向到真人的好时机。Off-task or "personal" questions directed to the agent are a good time to remind users of the synthetic nature of the agent and steer them to engage with it appropriately or to redirect them to a real person.

处理任务外问题

何时披露When to disclose

在整个用户旅程中,有很多进行披露的机会。There are many opportunities for disclosure throughout the user journey. 为第一次使用、第二次使用、第 n 次使用而设计,同时抓住“失败”的时机(如系统出错或用户发现代理功能的限制时)来强调透明度。Design for the first use, second use, nth use…, but also embrace moments of "failure" to highlight transparency—like when the system makes a mistake or when the user discovers a limitation of the agent's capabilities.

整个用户旅程中的披露机会

突出各种披露机会的标准数字助理用户旅程示例。Example of a standard digital assistant user journey highlighting various disclosure opportunities.

预先披露Up-front

最佳披露时刻是用户第一次与合成语音进行交互时。The optimal moment for disclosure is the first time a person interacts with the synthetic voice.  在个人语音助理方案中,应该是载入过程中或用户首次实际开启体验时。  In a personal voice assistant scenario, this would be during onboarding, or the first time the user virtually unboxes the experience. 在其他方案中,可能是合成语音首次朗读网站上的内容时,或者用户首次与虚拟角色进行交互时。In other scenarios, it could be the first time a synthetic voice reads content on a website or the first time a user interacts with a virtual character.

在请求时披露Upon request

用户应该能在发起请求后轻松访问额外信息、控制首选项,并可在用户旅程中的任何时候接收透明讯息。Users should be able to easily access additional information, control preferences, and receive transparent communication at any point during the user journey when requested.

持续披露Continuously

使用可持续增强用户体验的隐式设计模式。Use the implicit design patterns that enhance the user experience continuously.

在系统故障时披露When the system fails

使用披露作为从容应对故障的机会。Use disclosure as an opportunity to fail gracefully.

其他资源Additional resources

参考文档Reference docs

后续步骤Next steps