Privacy Policy

Visitors

Visits are recorded by the site itself, and also counted with Umami and Vercel Web Analytics: page views per page and day, approximate location (country, region, city), browser and device type, the referring site, and clicks on paper links. No cookies are used and no IP addresses are stored; visitors are not identified, not linked from one day to the next, and not followed across other sites. The site's own statistics are kept with the site's data for as long as the site runs; Umami keeps its statistics on servers in the EU. Totals may be shown on the site.

The public pages set no cookies. When you play a video in the gallery, the policy of the service that hosts it (Google Drive, YouTube, …) applies from then on.

Ontology search

Questions entered on the ontology page are sent to the OpenAI API only when a large language model (LLM) is used to interpret them. The question text is the only visitor-record data sent; the ontology vocabulary is also supplied as instructions. Requests are handled by keyword search without being sent to OpenAI when the daily or monthly usage limit is reached, the API key is absent, question logging cannot be verified, or the request is identified as coming from a crawler or script. OpenAI does not use API content to train its models by default and may retain it for a limited period to prevent abuse.

To improve search and limit excessive requests, the site records the question text (up to 300 characters), time, language, approximate location (country, region and city), and a visitor identifier. After the search, it also records how the question was answered (LLM interpretation or keyword search), the number of papers found, and the interpreted search conditions. IP addresses themselves are not stored; visitor identifiers are generated using a key that changes daily. Questions from requests identified as coming from crawlers or scripts are not recorded. Question records are automatically deleted after one year. Please do not enter personal information in the question box.

When you press "Copy link" under an answer, the site stores the question text, the conditions taken out, the language and the time, to make a short link; nothing about the visitor is stored. These records are kept for as long as the site runs, so that shared links keep working.

Personal information on the site

Names, photos, email addresses and researcher profiles (Google Scholar, ORCID, …) of members and alumni are shown with their consent, to introduce the lab. To have yours changed or removed, contact us below.

Author names and affiliations of papers and conference presentations come from works that are already published.

Administrators

For the lab members who maintain the site, we keep their accounts (email address, encrypted password) and a record of who changed what and when, for administration only. Cookies that keep them signed in are used on the admin pages, and on the public pages only to leave the lab members’ own visits out of the visit statistics.

Storage and processors

The site runs on Vercel (hosting) and Supabase (database and file storage); data may be stored on servers outside Korea (e.g. in the United States).

The site's data is backed up weekly to GitHub (kept 90 days) and to the PI's Google Drive (latest 12 weeks); older copies are deleted automatically.

Personal information is never sold or used for advertising, and is not given to third parties except where the law requires it.

Contact

Questions, and requests to see, correct or delete personal information: Prof. Byung-Hyun Kim (bhkim00 at hanyang.ac.kr)