版本：v1.0 生效日期：2026-11-08 运营主体：灵枢（南阳）智能科技有限公司（下称"我们"） 联系邮箱：contact@ziyoushikong.cn
《图解深度学习》（Deep Learning Lab）不收集、不上传、不共享任何个人信息。你的学习进度、收藏、语言偏好等全部只保存在你自己设备的本地存储中，只有你（以及设备备份机制）能接触到它们。卸载应用即彻底删除这些数据。
本政策适用于 macOS 与 iOS（iPhone / iPad）版本，以及其他平台的同类客户端。
我们不要求注册账号，不收集姓名、手机号、邮箱、地址，不采集设备标识符（IDFA / IDFV / 广告标识）、不采集位置、不采集通讯录 / 相机 / 麦克风 / 照片库 / 健康数据，也不接入任何广告、统计或归因 SDK。应用的核心功能在完全离线状态下即可使用。
| 存储项 | 内容 | 用途 |
|---|---|---|
| 会员解锁标记 | 是否已解锁的布尔标记 | 决定付费内容是否可见 |
| 已完成课程 | 已通关课程的标识集合 | 首页「继续学习」、进度统计 |
| 最近一节课程 | 单个课程标识 | 断点续学 |
| 连续学习天数 / 最近学习日期 | 天数与日期（YYYY-MM-DD） | 「连续学习」激励展示 |
| 任务关卡星级 | 关卡编号与星级 | 任务页展示与最优成绩 |
| 奖杯与完课里程碑 | 已颁发里程碑的标识与颁发日期 | 奖杯陈列与生成完课证明 |
| 术语收藏 | 你收藏的术语名称列表 | 术语页收藏功能 |
| Playground 参数 | 你在实验里调出的参数（JSON 串） | 下次打开时恢复实验现场 |
| 界面语言偏好 | 系统 / 中文 / 英文 | 界面语言跟随 |
| 本地使用计数 | 少数功能入口的被点次数（如付费页展示、解锁成功） | 仅用于我们判断功能改进，只存本机、不上传、不与任何个人身份关联 |
上述数据以键值形式存放于操作系统的本地偏好存储（iOS / macOS 的 UserDefaults），受系统沙盒保护。除你主动使用"分享"功能外，它们不会离开这台设备。
应用内的会员购买通过 Apple App Store 内购完成。支付金额、支付凭证、账单与身份信息全部由 Apple 处理，我们不接收、不查看、也不存储你的银行卡号或支付账号。 我们只在本机记录"已解锁"这一状态；订单与退款事宜请按 Apple 的退款政策在 App Store 中申请。
若后续版本引入云端能力（跨设备同步、账号体系等），我们将在上线前更新本政策，并在应用内重新征求你的同意。
为便于核查起见，我们列出应用内嵌的开源运行时与组件，它们均不收集或上传个人数据：Flutter 运行时（BSD）、shared_preferences（本地存储读写）、share_plus（唤起系统分享面板）、url_launcher（唤起系统浏览器）。此外应用内的可视化示例借鉴了 TensorFlow Playground 等开源项目，授权信息展示在应用"关于我们"页。
本产品定位为面向高中及以上学习者与成年自学者的教育工具，不专门面向 13 周岁以下儿童，不为其设计内容或进行投放。我们不收集任何用户的个人信息，因此也不存在收集未成年人个人信息的情形。
若你是未成年人的监护人，发现设备上的应用数据需要处理，可通过本文联系邮箱要求我们协助（在零收集架构下，通常的处理方式就是在设备本地删除）。我们建议未成年人在监护人指导下使用应用内的外链浏览功能。
零收集架构本身就是最强的安全措施：没有服务端数据库，就不存在服务端泄露。此外：应用运行在操作系统沙盒内，不绕过沙盒限制，本地存储不写入任何敏感信息；网络能力仅在你主动点击外链时由系统浏览器承担。若未来接入服务端，我们将采用传输加密（TLS）、最小化字段与访问控制，并重新评估本政策。
政策更新会在本页面公布，并在应用内以显著方式提示（重大变更会重新征求同意）。我们建议教育机构和教师在使用本应用布置教学任务前，将本政策链接一并纳入课程说明。
隐私相关问题（包括投诉与监管问询）请发送至：contact@ziyoushikong.cn，或邮寄至河南省南阳市社旗县建设北路东侧8号 473300。我们将在 15 个工作日内答复。
Version: v1.0 · Effective date: 2026-11-08 · Operator: Lingshu (Nanyang) Intelligent Technology Co., Ltd. ("we", "us") · Contact: contact@ziyoushikong.cn
Deep Learning Lab collects, uploads and shares no personal data at all. Your lesson progress, favourites and language preference live only in local storage on your own device; nothing about you exists on our servers, because there is no server. Deleting the app deletes the data.
This policy covers the macOS and iOS (iPhone / iPad) apps and any future build of the same product.
No account or sign-up is required. We do not collect your name, e-mail, phone number, address, photos, contacts or location. We do not collect device identifiers (IDFA, IDFV, advertising identifiers), we do not run any advertising, attribution or analytics SDK, and we do not request any privacy permission (location, camera, microphone, photo library, contacts, health). All core functionality works fully offline.
| Item | Content | Purpose |
|---|---|---|
| Membership flag | whether premium content is unlocked | gating paid lessons |
| Completed lessons | set of lesson identifiers | "Continue learning", progress stats |
| Last lesson | one lesson identifier | resume where you left off |
| Study streak / last study date | day count and a YYYY-MM-DD date | streak display |
| Mission stars | level id and star rating | best score per mission |
| Trophy milestones | milestone codes and the dates they were awarded | trophy shelf and completion-card rendering |
| Favourite terms | list of term names you starred | glossary favourites |
| Playground settings | the parameters you dialled in (JSON) | restore your experiment |
| Language preference | system / Chinese / English | UI language |
| Local counters | how often a few screens were opened (e.g. paywall shown, unlock succeeded) | our own product judgement only; stored on-device, never uploaded, never linked to a person |
These are plain key-value entries in the operating system's local preference store (UserDefaults on iOS / macOS), protected by the app sandbox. They leave the device only if you yourself use the Share feature.
Purchases are completed by Apple through App Store in-app purchase. Apple processes the payment, the receipt, the billing address and your identity; we never receive, view or store your card details or payment account. We only remember locally that content is unlocked. Refunds are handled by Apple under its own refund policy — request them in App Store / through reportaproblem.apple.com.
If a later version introduces server-side features (accounts, cross-device sync), we will update this policy before launch and ask for your consent again in the app.
For transparency: the app embeds open-source runtime components, none of which collect or transmit personal data — the Flutter engine (BSD), shared_preferences (local key-value storage), share_plus (system share sheet), url_launcher (system browser). Visualisations inspired by open-source projects such as TensorFlow Playground are credited on the in-app "About" screen.
The product is designed for high-school-level and adult self-learners and is not directed at children under 13; we do not design content for or advertise to that audience. Because we collect no personal information from anyone, we likewise collect none from children.
If you are a parent or guardian of a minor and need something handled regarding this app on a device, contact the address below — in a zero-collection architecture the remedy is normally deleting the data on the device itself. We suggest that minors use the external-link feature only with guidance.
Zero collection is the strongest control available: with no backend database there is no backend to breach. The app runs sandboxed, does not circumvent sandbox restrictions, stores nothing sensitive locally, and reaches the network only through your own browser when you tap an external link. Should a server ever be introduced, we will use TLS in transit, minimise fields and apply access control, and reassess this policy.
Updates are published on this page and surfaced in the app; material changes will be re-consented in-app. Schools and teachers assigning work with this app are encouraged to link this policy from their course materials.
Privacy questions, complaints and regulator enquiries: contact@ziyoushikong.cn, or mail to 河南省南阳市社旗县建设北路东侧8号 473300. We answer within 15 working days.
图解深度学习 · Deep Learning Lab