On Tue, Oct 10, 2017 at 7:31 AM, Andreas Kolbe jayen466@gmail.com wrote:
Wikidata has its own problems in that regard that have triggered ongoing discussions and concerns on the English Wikipedia.[1]
Tensions between different communities with overlapping but non-identical objectives are unavoidable. Repository projects like Wikidata and Wikimedia Commons provide huge payoff: they dramatically reduce duplication of effort, enable small language communities to benefit from the work done internationally, and can tackle a more expansive scope than the immediate needs of existing projects. A few examples include:
- Wiki Loves Monuments, recognized as the world's largest photo competition - Partnerships with countless galleries, libraries, archives, and museums - Wikidata initiatives like mySociety's "Everypolitician" project or Gene Wiki
This is not without its costs, however. Differing policies, levels of maturity, and social expectations will always fuel some level of conflict, and the repository approach creates huge usability challenges. The latter is also true for internal wiki features like templates, which shift information out of the article space, disempowering users who no longer understand how the whole is constructed from its parts.
I would call these usability and "legibility" issues the single biggest challenge in the development of Wikidata, Structured Data for Commons, and other repository functionality. Much related work has already been done or is ticketed in Phabricator, such as the effective propagation of changes into watchlists, article histories, and notifications. Much more will need to follow.
With regard to the issue of citations, it's worth noting that it's already possible to _conditionally_ load data from Wikidata, excluding information that is unsourced or only sourced circularly (i.e. to Wikipedia itself). [1] Template invocations can also override values provided by Wikidata, for example, if there is a source, but it is not considered reliable by the standards of a specific project.
If a digital voice assistant propagates a Wikimedia mistake without telling users where it got its information from, then there is not even a feedback form. Editability is of no help at all if people can't find the source.
I'm in favor of always indicating at least provenance (something like "Here's a quote from Wikipedia:"), even for short excerpts, and I certainly think WMF and chapters can advocate for this practice. However, where short excerpts are concerned, it's not at all clear that there is a _legal_ issue here, and that full compliance with all requirements of the license is a reasonable "ask".
Bing's search result page manages a decent compromise, I think: it shows excerpts from Wikipedia clearly labeled as such, and it links to the CC-BY-SA license if you expand the excerpt, e.g.: https://www.bing.com/search?q=france
I know that over the years, many efforts have been undertaken to document best practices for re-use, ranging from local community-created pages to chapter guides and tools like the "Lizenzhinweisgenerator". I don't know what the best-available of these is nowadays, but if none exists, it might be a good idea to develop a new, comprehensive guide that takes into account voice applications, tabular data, and so on.
Such a guide would ideally not just be written from a license compliance perspective, but also include recommendations, e.g., on how to best indicate provenance, distinguishing "here's what you must do" from "here's what we recommend".
Wikidata will often provide a shallow first level of information about a subject, while other linked sources provide deeper information. The more structured the information, the easier it becomes to validate in an automatic fashion that, for example, the subset of country population time series data represented in Wikidata is an accurate representation of the source material. Even when a large source dataset is mirrored by Wikimedia (for low-latency visualization, say), you can hash it, digitally sign it, and restrict modifiability of copies.
Interesting, though I'm not aware of that being done at present.
At present, Wikidata allows users to model constraints on internal data validity. These constraints are used for regularly generated database reports as well as on-demand lookup via https://www.wikidata.org/wiki/Special:ConstraintReport . This kicks in, for example, if you put in an insane number in a population field, or mark a country as female.
There is a project underway to also validate against external sources; see:
https://www.mediawiki.org/wiki/Wikibase_Quality_Extensions#Special_Page_Cros...
Wikidata still tends to deal with relatively small amounts of data; a highly annotated item like Germany (Q183), for example, comes in at under 1MB in uncompressed JSON form. Time series data like GDP is often included only for a single point in time, or for a subset of the available data. The relatively new "Data:" namespace on Commons exists to store raw datasets; this is only used to a very limited extent so far, but there are some examples of how such data can be visualized, e.g.:
https://en.wikipedia.org/wiki/Template:Graph:Population_history
Giving volunteers more powerful tools to select and visualize data while automating much of the effort of maintaining data integrity seems like an achievable and strategic goal, and as these examples show, some building blocks for this are already in place.
But the proprietary knowledge graphs are valuable to users in ways that the previous generation of search engines was not. Interacting with a device like you would with a human being ("Alexa/Google/Siri, is yarrow edible?") makes knowledge more accessible and usable, including to people who have difficulty reading long texts, or who are not literate at all. In this sense I don't think WMF should ever find itself in the position to argue _against_ inclusion of information from Wikimedia projects in these applications.
There is a distinct likelihood that they will make reading Wikipedia articles progressively obsolete, just like the availability of Googling has dissuaded many people from sitting down and reading a book.
There is an important distinction between "lookup" and "learning"; the former is a transactional activity ("Is this country part of the Euro zone?") and the latter an immersive one ("How did the EU come about?"). Where we now get instant answers from home assistants or search engines, we may have previously skimmed, or performed our own highly optimized search in the local knowledge repository called a "bookshelf".
In other words, even if some instant answers lead to a drop in Wikipedia views, it would be unreasonable to assume that those views were "reads" rather than "skims". When you're on a purely transactional journey, you appreciate almost anything that shortens it.
I don't think Wikimedia should fight the gravity of a user's intentions out of its own pedagogical motives. Rather, it should make both lookup and learning as appealing as possible. Doing well in the "lookup" category is important to avoid handing too much control off to gatekeepers, and being good in the "learning" category holds the greatest promise for lasting positive impact.
As for the larger social issue, at least in the US, the youngest (most googley) generation is the one that reads the most books, and income/education are very strong predictors of whether people do or not: http://www.pewresearch.org/fact-tank/2015/10/19/slightly-fewer-americans-are...
The applications themselves are not the problem; the centralized gatekeeper control is. Knowledge as an open service (and network) is actually the solution to that root problem. It's how we weaken and perhaps even break the control of the gatekeepers. Your critique seems to boil down to "Let's ask Google for more crumbs". In spite of all your anti-corporate social justice rhetoric, that seems to be the path to developing a one-sided dependency relationship.
I considered that, but in the end felt that given the extent to which Google profited from volunteers' work, it wasn't an unfair ask.
While I think your proposal to ask Google to share access to resources it already has digitized or licensed is worth considering, I would suggest being very careful about the long term implications of any such agreements. Having a single corporation control volunteers' access to proprietary resources means that such access can also be used as leverage down the road, or abruptly be taken away for other reasons.
I think it would be more interesting to spin off the existing "Wikipedia Library" into its own international organization (or home it with an existing one), tasked with giving free knowledge contributors (including potentially to other free knowledge projects like OSM) access to proprietary resources, and pursuing public and private funding of its own. The development of many relationships may take longer, but it is more sustainable in the long run. Moreover, it has the potential to lead to powerful collaborations with existing public/nonprofit digitization and preservation efforts.
Publicise the fact that Google and others profit from volunteer work, and give very little back. The world could do with more articles like this:
https://www.washingtonpost.com/news/the-intersect/wp/2015/07/22/you-dont-kno...
I have plenty of criticisms of Facebook, but the fact that users don't get paid for posting selfies isn't one of them. My thoughts on how the free culture movement (not limited to Wikipedia) should interface with the for-profit sector are as follows, FWIW:
1) Demand appropriate levels of taxation on private profits, [2] sufficient investments in public education and cultural institutions, and "open licensing" requirements on government contracts with private corporations.
2) Require compliance with free licenses, first gently, then more firmly. This is a game of diminishing returns, and it's most useful to go after the most blatant and problematic cases. As noted above, "fair use" limits should be understood and taken into consideration.
3) Encourage corporations to be "good citizens" of the free culture world, whether it's through indicating provenance beyond what's legally required, or by contributing directly (open source development, knowledge/data donations, in-kind goods/services, financial contributions). The payoff for them is goodwill and a thriving (i.e. also profitable) open Internet that more people in more places use for more things.
4) Build community-driven, open, nonprofit alternatives to out-of-control corporate quasi-monopolies. As far as proprietary knowledge graphs are concerned, I will reiterate: open data is the solution, not the problem.
Cheers, Erik
[1] See the getValue function in https://en.wikipedia.org/wiki/Module:WikidataIB , specifically its "onlysourced" parameter. The module also adds a convenient "Edit this on Wikidata" link to each claim included from there.
[2] As far as Wikimedia organizations are concerned, specific tax policy will likely always be out of scope of political advocacy, but the other points need not be.