Wikimedia Deutschland/Data Partnerships Model/Data Partnership Scenarios

From Meta, a Wikimedia project coordination wiki
Jump to navigation Jump to search
Data Partnerships Model Banner.png


 Main page Executive Summary Background Purpose Partners The Phases of Data Partnerships Data Partnerships Scenarios Decision-Making Conclusion 
Data partnership scenarios

There are several scenarios that are most common in our current data partnerships work. The following list covers those cases and we may add to it in the future.

1. Structured data donation[edit]

Type of partners: national libraries, museums and archives of high relevance, international organizations, research institutions and government organisations.

Contact: they typically contact us.

How does the project help them? There are many reasons why a data donation might help a partner, including but not limited to increasing visibility, enrichment, multilingual support and revealing potential biases in their collection.

How does the project help us? Wikidata gets more access to high-quality data.

What is our role? We provide information on how to make a data donation (presentation, answering questions via email etc.) and initiate contact between partners and community members who may help with uploading the donation. We support the donation by publicizing it (e.g., in a blogpost).

Do we provide technical support? No.

Who decides if we take on this partner? The community makes this decision. Where possible, we will connect the partner to a community member willing and able to aid the donation process. However, it is up to the community to decide whether to accept the donation.

Questions to ask in this scenario:

  • Is the data notable according to Wikidata community rules?
  • Does the donation consist of high-quality data?
  • Regarding the scope of the improved data: how big is the set?
  • Does the data fall within Wikidata’s scope?
  • Will a community member or in-house expert assist in both uploading and maintaining the data on an ongoing basis?

This is the most obvious and, historically, the first kind of data partnership: a partner has a set of data that they want to donate to Wikidata, creating new Wikidata Items. GLAM projects have focused on this kind of partnership, which provides visibility for the institution’s data (benefiting them) and more quality content for our projects (benefiting us). Data donations are usually handled by the community after our initial contact. That process is documented here. In this case, communications with the partner should stress that the partner making the donation becomes part of the community and does not “own” the Items, since of course anything may be edited by anyone. The signed statements feature in Wikidata, planned for the future, may address some of the concerns partners may have about ownership. This feature will increase the visibility of partnership contributions and may very well influence these kinds of partnerships.

2. Linking to data[edit]

Type of partners: organizations having some kind of catalogue with persistent identifiers (e.g., libraries, companies like Quora, museums).

Contact: they typically contact us.

How does the project help them? Their data gets more visibility and is made available via Wikidata. Their data gets linked to other identifiers, making their identifiers discoverable and creating linked data that connects realms (e.g., a bibliographic identifier linked to GND, VIAF, or LOC).

How does the project help us? Wikidata strengthens its role as a hub in the Linked Open Data web. Wikidata gets more access to high-quality data.

What is our role? We provide information on how to make a data donation (presentation, answering questions via email, etc.) and initiate contact with community members who may help with uploading the donation. We support the donation by publicizing it (e.g., in a blogpost).

Do we provide technical support? No.

Who decides if we take on this partner? The community makes this decision. Where possible, we will connect the partner to a community member willing and able to aid the donation process. However, it is up to the community to decide whether to accept the donation.

Questions to ask in this scenario:

Linking to data can be seen as a special case of data donation. An institution’s data sets may contain identifiers that can be linked to Q-IDs in Wikidata, thus strengthening Wikidata’s role as a hub in the Linked Open Data Web. A good example case — though not actually a Wikimedia Deutschland partnership — is the Library of Congress.

3. Enriching data[edit]

Type of partners: institutions that are experts in a field or have data that is considered authoritative (e.g., the German Integrated Authority File).

Contact: they typically contact us.

How does the project help them? The partner’s data gets more visibility and is made available on Wikidata. They strengthen their reputation as an authoritative source.

How does the project help us? Wikidata gets more access to high-quality data.

What is our role? We provide information on the process of data donation (presentation, answering questions via email etc.) and initiate contact with community members who may help with uploading the donation. We support the donation by publicizing it (e.g., in a blogpost).

Do we provide technical support? No.

Who decides if we take on this partner? The community makes this decision. Where possible, we will connect the partner to a community member willing and able to aid the donation process. However, it is up to the community to decide whether to accept the donation.

Questions to ask in this scenario:

  • Is the data notable according to Wikidata community rules?
  • Does the donation consist of high-quality data?
  • Does the data fall within Wikidata’s scope?
  • Regarding the scope of improved data: how big is the set?
  • Will a community member or in-house expert assist in both uploading and maintaining the data on an ongoing basis?
  • Are there existing identifiers that can be used to reconcile the new data set with Wikidata?

Enriching data is another special form of data donation. An institution may be an expert or an authority in a field and may work to add new statements to existing Wikidata Items, thereby enriching the data. As with scenario one, communications with the partner should stress that the partner enriching the data becomes part of the community and does not “own” the Items, since of course anything may be edited by anyone. The signed statements feature in Wikidata, planned for the future, may address some of the concerns partners may have about ownership. This feature will increase the visibility of partnership contributions and may very well influence these kinds of partnerships.

4. Installing a Wikibase instance that may or may not link to Wikidata[edit]

Type of partners: partners having structured data who want to use Wikibase for its features (often libraries with authority files, but also other GLAMs or institutions of any kind)

Contact: they typically contact us.

How does the project help them? They can explore and evaluate Wikibase for their structured data needs. In many cases, Wikibase is the only solution to flexibly model, collaboratively edit and query linked data.

How does the project help us?

  • More knowledge is made accessible to more people
  • Increased insight into the user requirements for Wikibase
  • Increased visibility of Wikibase
  • The Wikibase ecosystem is strengthened

What is our role? We provide information on the installation (presentation, workshops, answering questions via email, etc.). We support the installation by publicizing it (e.g., in a blogpost). We may also invest time and other resources (development, UX).

Do we provide technical support? Every partnership is different; the answer here can vary from yes to no.

What happens if we decide to take this partnership opportunity? We apply our decision-making process, which determines whether we accept the partnership. The criteria for this process is outlined in the decision-making section of this document.

Who decides if we take on this partner? Wikimedia Deutschland.

Questions to ask in this scenario:

  • Is Wikibase suitable for this data set?
  • Does the potential partner have the technical expertise to install, maintain and host Wikibase?
  • Does the potential partner have the technical expertise to model, upload and maintain the data?

We actively explore this option for institutions that have structured data which may not be within Wikidata’s scope. The Gemeinsame Normdatei (GND) Wikibase evaluation at the German National Library is a good example; interest in the library world is very high. For institutions, especially libraries, Wikibase is often the most attractive open-source solution for structured and linked data. Partnerships such as these allow us to collect needs and requirements to aid our development process and demonstrate how Wikibase works in the wild. Additionally, such projects show that Wikibase can successfully power non-Wikimedia projects approaching the scale of Wikidata.

5. Hybrid[edit]

Partners vary, and so may their needs and the data they can provide. Partnerships may also be a mix of some or all of the scenarios mentioned above, or they may start as one scenario and develop into another. One example of this is our partnership with the German National Library, mostly a project to help with the evaluation of Wikibase (scenario 4), but it may also have the side effect of enriching Wikidata Items with updated GND numbers (scenario 3).