# Maintaining data ## What maintaining data means Maintaining data on HDX is the process to ensure datasets are kept accurate and up-to-date throughout their lifecycle. This includes updating files when new information becomes available, correcting or improving metadata, monitoring alerts from HDX, responding to requests for underlying data, and making sure each dataset has a person assigned to oversee its quality and relevance (referred to as a ‘data maintainer’). ## Why maintaining data matters Maintaining your data on HDX is essential to ensure it remains useful, accurate and reliable for humanitarians. By regularly reviewing and updating your data you can make sure that HDX remains a source for reliable and high-quality humanitarian data. This section explains the following key tools and processes that help you manage and maintain your data on HDX: * The User Dashboard and My Datasets tab * Understanding alerts through the notifications bell * How to edit or update a dataset * The role of data maintainers and how to manage them * How to respond to HDX Connect data requests * How and why datasets are archived on HDX ## The User Dashboard The User Dashboard is your central workspace on HDX for managing and maintaining data. It brings together the datasets you maintain, your organization memberships, data requests, and updates from across the platform. To access the User Dashboard, click on your name in the top banner of any page.

**Newsfeed** The Newsfeed displays recent activity across HDX, including updates to datasets you follow and changes made by HDX system bots. HDX bots perform automated background checks, maintenance and updates, such as metadata validation, to help maintain consistency and data quality across the platform. **My Datasets** To view the My Datasets tab click on your name in the top banner and scroll down to User Dashboard. The My Datasets tab shows all datasets that you maintain on HDX, and their update status. From this page, you can quickly see which datasets are up to date, which need attention, and which have outdated metadata. Filters on the left help you narrow results by update status, location, or file format. Each dataset listed displays information such as: * Time period of the dataset * Expected update frequency and warning if the expected update date has passed * Links to edit or delete the dataset

**HDX Connect Requests** The HDX Connect Requests tab is where you manage all HDX Connect (by request) data access requests. Learn more about [HDX Connect](/publish/publish-data/preparing-and-publishing-data.md#controlled-access-publishing). There are three sections: * **New:** Requests you have not yet responded to * **Open:** Requests you have replied to but not archived * **Archived:** Previously completed or declined requests From here you can reply, decline, track your response to previous requests, and/or reopen past requests.

## The Notifications Bell The notifications bell in the top header of HDX alerts you to important actions that require your attention.

When a red dot appears, it means you have new notifications related to: 1. Datasets needing updates based on their expected update frequency 2. Pending HDX Connect requests for datasets you maintain 3. Quality Assurance updates, such as datasets placed Under Review 4. Messages or activity related to your organization’s datasets Click the bell to open the notifications panel, where you can view all alerts and sort as needed. Take action by navigating directly to the pages. This feature helps data maintainers to quickly respond to requests, keep their data up-to-date, and stay informed about QA-related actions. ## How to edit or update a dataset on HDX HDX will notify you when a dataset you maintain needs to be updated. To review these alerts, click the notifications bell icon in the top header. An example notification is shown below, where five datasets need to be updated. Please note the notifications bell is not connected to the notifications feature. Learn about the [notifications feature](/find/find-data/staying-up-to-date-with-hdx.md#hdx-notifications). **Common reasons to update your dataset on HDX** * The data is out-of-date, one of the most frequent triggers for updates * There are new resources to add * The metadata needs correcting * The sharing level needs to change, ex. public to private ### **How to manually edit a dataset on HDX** The Edit button appears just below your organization’s logo on the dataset page. {% stepper %} {% step %} #### Click Edit

You can only edit a dataset if you are an [administrator or editor](/publish/publish-data/organizations-on-hdx.md#membership-roles-and-permissions) of the HDX organization. If you are logged in, and have the appropriate role,you will find an ‘Edit’ button just below the organization logo. {% endstep %} {% step %} #### Make changes in the Edit data form

The edit form contains the same details in the [share data form](/publish/publish-data/preparing-and-publishing-data.md#publishing-data-on-hdx-manually-how-to-use-the-share-data-form). Common edits include: * Updating the title or description * Updating the time period or adding files when new data becomes available * Correcting or expanding metadata fields * Assigning a new data maintainer if responsibilities have changed * Changing the sharing level (e.g., private, public, HDX Connect) after reassessing sensitivity or internal policy * Replacing links to remote files that have moved or expired (example: Google Drive) {% endstep %} {% step %} #### Check the tick box to confirm the dataset does not contain Personally Identifiable Data (PII) or violate the HDX Terms of Service.

\[Important!] You must check the data sensitivity box before submitting your dataset. This confirms that you have reviewed the data and that it is safe to publish on HDX. Learn more about [assessing sensitive data here](/publish/publish-data/preparing-and-publishing-data.md#is-the-data-safe-to-publish). {% endstep %} {% step %} #### Click Submit Dataset If your dataset is updated programmatically through a pipeline, refer to our Github repository to learn more about the Python functions: {% endstep %} {% endstepper %} ### Understanding dataset maintainers Every dataset on HDX must have a designated data maintainer. They provide a point of contact for users, receive update reminders, and are responsible for keeping the dataset up-to-date. It is important that organizations review their data maintainers, ensuring continuity as teams change. Without an assigned maintainer, requests may go unanswered and datasets can quickly become outdated. Dataset maintainer role: * Update datasets when new versions are available. * Respond to user questions via “Contact the Contributor.” * Receive HDX Connect requests and decide whether data can be shared. * Ensure responsible data sharing, including sensitivity checks and metadata accuracy. **How to reassign a data maintainer** If a team member leaves an organization (or moves roles) their datasets must be reassigned to a new data maintainer. To reassign a maintainer: 1. Open the dataset 2. Click Edit 3. Scroll to the Maintainer field 4. Select a new person from the dropdown list (must be a member of the organization on HDX) 5. Save changes by clicking Submit Dataset Maintaining an active data maintainer is required to ensure accountability and continuity. If a new maintainer is not assigned, the HDX team will contact the organization’s administrators to flag the issue and request that one be added. If a new data maintainer cannot be assigned after follow-up, the dataset may be archived to maintain data quality and governance standards on the platform. ## Managing HDX Connect requests When a metadata-only dataset is published using HDX Connect, registered HDX users can request access to the underlying data. The dataset maintainer is responsible for reviewing and responding to these requests through the organization dashboard. ### How to review and respond to user requests {% stepper %} {% step %} #### Review the request You will receive an email notification when someone requests access to one of your HDX Connect datasets. The email will include: * Requester’s name, email, location, and organization * The dataset name * The requester’s intended use * Their message explaining why they need the data To respond: * Click the “My Requests” link in the email OR * Click the red notification dot next to your username in the HDX header {% endstep %} {% step %} #### Manage requests on the My Requests page The My Requests page displays all requests for the data on HDX you maintain. The page has three sections: * **New**: New, unanswered requests that require your action. * **Open**: Requests you have replied to, but have not yet archived. The request will remain “Open” until you close the ticket and it is archived. Please close all open tickets once action is taken to ensure that we can effectively track requests. * **Archived**: Requests you declined or previously closed. You can reopen archived requests if needed. For each request, you will see the dataset name, the requester’s user details, and their reason for requesting the data, alongside ‘reply’ or ‘decline’ buttons. {% endstep %} {% step %} #### Respond to the request As the data maintainer, you have two options. You can either reply (to share the data or ask follow-up questions) or decline the request #### **If the data maintainer chooses to share the data or discuss further** Please note that HDX does not host or share the requested data. The data is shared directly between your organization and the requester, outside of the HDX platform. Sensitive data should always be shared through a secure channel. Share passwords separately, using a different secure channel. For more on [responsible approaches to data sharing](https://centre.humdata.org/data-responsibility/). 1. **Click Reply to open the response form.** In your reply: * Explain how you will share the data (e.g., encrypted email, secure transfer) * Ask any clarification questions if needed * Include any instructions the requester must follow 2. **Click Submit when finished.** The request will stay in Open status until you manually archive it. #### If you choose to decline the request 1. **Click Decline.** You will be prompted to: * Provide a brief explanation of why the data cannot be shared Once submitted: * The request is automatically moved to Archived HDX tracks whether a request is approved or declined, but does not track the content of your messages. {% endstep %} {% step %} #### Keep track of shared data In the Open section, HDX displays Yes/No buttons to track whether you have shared the data\ with the requester. This helps your organization monitor its data-sharing actions over time. {% endstep %} {% endstepper %} ## Archiving data on HDX To keep HDX a reliable and up‑to‑date resource for humanitarian data, older datasets that are no longer operationally relevant are periodically archived by the HDX team. Archiving ensures users can more easily find current data while still offering access to older datasets for research or historical analysis. Archived datasets remain available and searchable but are moved under a separate Archived Datasets toggle and marked with a folder icon. They are excluded from the main dataset listings by default.

**When and how are datasets archived?** The HDX team reviews datasets on an annual basis and applies both automated and manual checks to identify which should be archived. Archived datasets typically meet the following criteria: * Created more than five years ago * Reference period ended more than five years ago * Fewer than 1,000 total downloads * Not categorized as requestable (HDX Connect) Certain datasets and organizations are exempt from archiving, including [Common Operational Datasets (CODs)](https://data.humdata.org/cod), and organizations with historical data that are still operationally relevant such as World Bank development data. If a dataset is archived and later found to be relevant again, the owning organization can request to unarchive it at any time. --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://docs.humdata.org/publish/publish-data/maintaining-data.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.