# Publish data

## If your data is already online

If you have a link to published data, that's great! You can [**submit it to our database**](https://airtable.com/app473MWXVJVaD7Es/shrJafakrcmTxHU2i) where it will show up in searches. We'll trigger automatic archives, and people who subscribe to data in the area will be notified.

## If you need to publish data first

If you have public records or data you want to share more broadly, there are a few ways to go! Make sure to [**share your published data when you're done**](https://docs.pdap.io/activities/submit-data-sources).

### Options for sharing data

{% hint style="warning" %}
If this is overwhelming, don't worry—every case is different, and we can provide advice. Email <contact@pdap.io> for a free consultation.
{% endhint %}

* [GitHub](https://github.com/) is a free platform intended for storing open-source code. It can be a great choice for sharing a small dataset, especially if that comes with a Python notebook of analysis or a web scraper which generated the data. [Here is a demo](https://github.com/Police-Data-Accessibility-Project/github-actions-demo) of an automated web scraper which collects Oakland's Calls for Service data, for free.
* [JKAN](https://jkan.io/) (for smaller agencies) and [CKAN](https://ckan.org/) (for larger ones) are two open-source data portals which need to be set up by someone with some technical skills.
* DocumentCloud can be used for publishing PDFs, and has additional tools for recognizing text (OCR) and annotation. You will need to [request verification](https://airtable.com/shrZrgdmuOwW0ZLPM) to publish documents.
* Your town or region may have an open data portal ([here's an example from Pittsburgh area](http://www.wprdc.org/)) which may be interested in publishing the data you have.
* A service like [Dropbox](https://dropbox.com/) can be used to share folders of documents in bulk, with fewer tools for keeping them organized or labeled.

## Making published data accessible

Sometimes, published data is not really *accessible*—maybe the data is rendered as a visualization on a dashboard, or you want to calculate a sum or average value from a 250-page HTML table. That's where [web scraping](https://docs.pdap.io/activities/web-scraping) comes in.