Privacy Policy
Nihongo to Japan takes your privacy seriously. This policy explains how we collect, use, and protect your information (last updated: September 9, 2026).
Data Controller
This website (Nihongo to Japan, www.nihongotojapan.com) is operated by Lingyang Education Co., Ltd. (Taiwan unified business number 60455916; 6F-3, No. 1, Jinshan N. Rd., Zhongzheng Dist., Taipei, Taiwan). Lingyang Education Co., Ltd. is the data controller for the personal data described in this policy.
What Data We Collect
This website uses Google Analytics (GA4) to automatically collect anonymized usage data, including: pages viewed, time on site, device type, and country/region. We do not actively collect your name, email address, or any personally identifiable information unless you contact us directly.
Your Personal Learning Profile (stays on your device)
Your personal learning profile — the data that powers the Learning Assistant: level estimate, five-axis skill radar, weakness analysis, lesson completion, flashcard review schedule, and mistake log — is stored entirely in your browser on this device (localStorage). It is not uploaded to our servers, and no account, name, or email is required. Clearing your browser data deletes it; we hold no copy. Cross-device sync is optional and only backs up when you turn it on and enter a sync code. The backup is not encrypted; what protects it is the sync code itself. We have no name or email for you, so we cannot connect it to a real-world person, but the backup contains this device's anonymous learner id, so technically it can be matched to that device's answer records.
Anonymous Answer Statistics (uploaded, used to improve questions and recommendations)
Separate from the personal profile above: each time you answer a question, we also send one de-identified answer event to our servers, containing only the question ID and question type, whether you were correct, the option you chose, timestamp and response time, JLPT level, skill axis and knowledge point (including secondary knowledge points), whether you viewed a hint or explanation, whether this was your first time seeing the question, whether you re-answered after seeing the answer, attempt number, which part of the site the answer came from, a randomly generated device identifier, a session identifier scoped to a single visit, and the evidence class, evidence weight and data/engine version kept so the figures can be recomputed later; it also includes the raw knowledge-point tag recorded at the moment you answered (labels such as は or conj:te, so past events can be re-classified if the knowledge-point scheme changes), the estimates that both of our mastery models held for you immediately before you answered (including that estimate's confidence and coverage), the predicted probability that you would answer that particular question correctly together with the difficulty-table version in use, and — when an answer falls inside an internal comparison of question-selection methods — the experiment identifier and the group you were assigned to. It contains no name, no email, and no IP address, and cannot be traced back to you. It is used for four things: calibrating each question’s real difficulty and discrimination, detecting broken questions, improving the Learning Assistant’s recommendations, and presenting these answers back to other learners as aggregate statistics (for example how many learners have attempted a given question, how many got it right on their first try, how many chose each option, and which mistakes are common). Anything presented publicly is an aggregate figure: it does not show any individual learner’s identity, any single answer record, when an answer was given, or the device identifier, and a question’s statistics are withheld entirely when too few learners have attempted it. Never used for advertising, never sold or shared with third parties. The data is append-only with no fixed deletion schedule at present. You can opt out in one click in section 1c of our privacy policy — uploads stop immediately and any queue on your device is cleared.
Score Reports and Creator Referral Links
If you actively submit a JLPT or mock-exam score report in your Study Hub, that single report is uploaded anonymously to calibrate our level estimates; it contains no name or email. Additionally, if you arrive via a partner creator’s referral link (a URL of the form /go/…), we store that link’s campaign parameters in your browser and attach them to subsequent learning-milestone events sent to GA4, so we can tell which referrals actually help learners. These likewise contain no name, email, or IP address, and are not used to track you across other websites.
Cookies and Tracking Technologies
This site uses cookies and similar technologies to improve your browsing experience and to support tracking by Google Analytics (GA4) and Google AdSense. You may disable cookies through your browser settings, though some features may be affected as a result.
Google AdSense and Affiliate Marketing Disclosure
This site displays advertisements through Google AdSense and includes affiliate marketing links to Klook and KKday. If you complete a purchase through an affiliate link, we may receive a small commission. This does not affect the price you pay or the neutrality of our content.
Social Platform Data (Threads)
We run the Threads account @nihongotojapan as the public-facing channel for our learning community. When you leave a public comment under a post from that account, or mention that account in a public post, we retrieve that public content through the Threads API provided by Meta, including your Threads username, the text of your comment, the time it was posted, and the public permalink to it. We use this for exactly two purposes: writing and publishing a reply to the question you raised, and measuring the reach and engagement of our own posts so we can decide what learning content to publish next. We also retrieve performance figures for our own posts (views, likes, replies, reposts); those figures contain nothing that identifies an individual user. We do not obtain your Threads direct messages, email address, phone number, or any non-public information, and we do not use this data for advertising, sell it, or share it with third parties. This data is held in our internal operational database, used to avoid duplicate replies and to review the quality of our answers, and is retained for 24 months, after which it is deleted. If you would like the record of your comment deleted, email us and we will delete it within 30 days; you may also delete your comment on Threads yourself at any time, and once deleted we retain it no further. We use an AI service provided by Anthropic PBC (United States) to help draft replies, and the public comment text described above is sent to that service for processing. Access tokens are not provided to any third party.
When You Contact Us Directly
If you reach out to us via Instagram DM or any other channel (for example, to discuss a partnership or a study-in-Japan inquiry), any information you share will be used solely to respond to your request and will not be passed on to third parties.
Your Rights
You may disable cookies through your browser settings, opt out of analytics data collection by installing the GA4 opt-out browser add-on, or turn off personalized ads through your Google Ads settings.
Contact and Data Requests
For any privacy question, data request, or deletion request, email [email protected], or contact us on Instagram @nihongotojapan. We respond within 30 days. (Plain-text form of the address, for environments that cannot run JavaScript: nihongo2japan [at] gmail.com)