Instead of collecting only a title and URL, record the reference date, publication date, access date, units, and terms of use in one row. Includes a CSV template and checklist that need no code.
This translation was generated by AI. Check the code, units, and numbers against the original. Native-speaker review has not yet been completed for each language. 한국어
Who this is forBeginners who find and organize public data themselves for reports, market research, or assignments
When you find data to use in a report, record one row right then. Keeping the data title, provider, and detail URL together lets you open the same data again later. Before copying numbers, also write what the data describes, as of when, and in what unit.
source_ledger.csv in this article is a table for recording the context of data. It has no function to collect numbers automatically or judge whether data is current. No actual crawling, API calls, or account connections were made.
| Item | Meaning | Example or writing rule |
|---|---|---|
| reference_date | The reference date the data describes | 2026-03-31 (fictional example) |
| published_date | The day the data was published | 2026-04-10 (fictional example) |
| checked_date | The day you actually opened and checked the source | Blank until you open it |
| unit | The unit in which values are counted or measured | As written in the source, such as items, people, cases, KRW, or thousand KRW |
Do not assume that a portal's modified date is the same as the data's reference date. If the data covers a period, such as monthly or annual data, add the start and end dates and whether it is cumulative to scope_definition. If no reference date is shown, do not substitute today's date; leave it as needing confirmation.
| Column group | CSV columns | What to record |
|---|---|---|
| Record type | record_id, data_kind | Unique ID and whether the data is fictional or real |
| Source | source_title, publisher, source_url | Original title, provider, and detail URL |
| Dates | reference_date, published_date, checked_date | Reference date, publication date, actual access date |
| Interpretation | unit, scope_definition | Unit, target, region, and counting definition |
| Terms of use | license_url, license_note | Where the terms of use for the data are and what was checked |
| Progress | verification_status, follow_up | Check status and the next item to check |
In source_url, put the data's detail page rather than a search result address. In license_url, record where you confirmed the terms that actually apply. If you downloaded a file, also write the file name and version in scope_definition or follow_up.
The data detail screen of the Korean Public Data Portal shows items such as provider, modified date, and permitted scope of use. You can use the actual detail page in the reference documents to practice reading those item names separately. That page is not the source of the fictional examples.
| ID | Fictional data | Key items left blank | Check status |
|---|---|---|---|
| EX001 | Facilities in fictional region A | Source URL, access date, terms-of-use URL | Source not checked |
| EX002 | Received requests in fictional region B | Common unchecked items + reference date | Reference date needs checking |
| EX003 | Usage in fictional region C | Common unchecked items + unit | Unit needs checking |
None of the 3 example rows is marked as record check complete. A blank does not mean 0 or not applicable; it means it could not be checked. No column for real statistics was created, so the examples cannot be cited as evidence for numbers.
The final result is not an automatic report, but a source list that lets you find the source again and track the remaining check items.
CHECKLIST.txt in the download contains the same check procedure and recommended status values. Record check complete is a work status; it does not guarantee the accuracy or currency of the data, or permission for every purpose of use.
| Problem | How to fix it |
|---|---|
| The source has only the portal name | Keep the original title, provider, and detail URL together. |
| The reference date was overwritten with the access date | Restore the reference date to the data's point in time and put the actual access date in checked_date. |
| Only a modified date is shown but it was recorded as the publication date | Do not guess published_date; write in follow_up that only the modified date was checked. |
| Thousand KRW was copied as KRW | Keep the original unit, and if conversion is needed, keep separate calculation evidence. |
| Blanks were filled with 0 | Express values you could not check with a blank and a needs-checking status. |
| Dates and Korean text in the CSV changed | In the import, choose UTF-8 and text columns, then save and reopen. |
The template structure was checked with the csv, json, and datetime standard libraries in CPython 3.12.14 on Windows 11 (10.0.26200). Users can edit the CSV without Python.
Includes code, input data, and instructions. Extract the ZIP and read README.txt first.
Download example ZIPExample code, filenames, and input keys remain unchanged. Refer to the commands and checking steps in the translated article as well.
The explanations and examples were written for this site. See the official sources below for the related behavior and concepts.