Topic on Talk:Wikimedia Technology/Annual Plans/FY2019/CDP3: Knowledge Integrity

Why do the several outcomes, targets, and outputs associated with the top-line goal to "increase reach" discuss only linking to external resources?

Does that actually even increase reach? Isn't reach defined as the number of people who (can) find us, not the number of things people can find from us?

Shouldn't we, if we wish to work to satisfy the goal as stated, work to make sure that the external sources who re-use our content for their customers link back to us, so that the content can be further updated? Or is there a new definition of reach being used for this goal? James Salsman (talk) 11:30, 18 April 2018 (UTC)

Hi James. Increasing reach in this instance is primarily through building a web of interconnected, structured, linked data and relationships. We don't control when others link to us, but we can map out the entire 'graph' of who we link to and how those links are related to other data components and sources. We are also increasing reach in other ways, by empowering better research which reaches readers--and by better informing the public of Wikipedia's verifiability and citation practices, to increase their understanding and better inform their usage of the site. Cheers, Jake

Hi Jake, are there any sources which define "reach" as the number of outgoing external links instead of the number of incoming channels?

Do the Executive Director and the Board agree with this nonstandard definition? If so, would you please publish their directly quoted statements to that effect? Thanks. James Salsman (talk) 19:40, 19 April 2018 (UTC)

James, it seems like you may be misinterpreting what we mean (or we're miscommunicating it). Take link archiving for example. If those links are dead, readers cannot reach content we cite. Yes, it exists outside our site domain network, but it's definitely a benefit of readers to reach them with working urls that they can reach. In the context of mapping how we link to external sources, that map itself is a product that increases our reach, because citation graphs are traditionally proprietary and closed access. The link graph *is* the product, and a tool in itself. Not only that, but having the link graph makes it easier to find related sources and do better research, which ultimately reaches readers.

@Eloquence: what is your definition of "reach" in the context of Wikimedia metrics? James Salsman (talk) 14:26, 20 April 2018 (UTC)

I have proposed a resolution to this at .

Does it require additional budget, FTE resources, or training? James Salsman (talk) 15:03, 18 April 2018 (UTC)

Hi James, I have reverted your edit so we can first discuss it here. Issues of privacy, "fuzzing", compliance are definitely significant increases in the scope of this program and would involve not only additional technical capacity but investment of legal counsel as well. At the moment all issues concerning privacy should be addressed directly to privacy{{at}}wikimedia.org. If you would like to propose those measure, that is the place to start. Cheers, Jake

Jake, has anyone performed a cost-benefit analysis of subpoena processing costs under Congress's new amendments to the Section 203 safe harbor provisions? Do you agree that it would be good to consider how the subpoena processing cost burden changes under storing and not storing personally identifying information (not I am not suggesting refraining from storing cryptographic hashes of IP addresses such as would be necessary for checkusers to investigate sockpuppetry)? If so, would you please ask for the question to be considered by Legal? Thank you. James Salsman (talk) 19:44, 19 April 2018 (UTC)

I will certainly pass the question on to Legal. Thank you! Jake

Hi James, these questions are way beyond my area of expertise and are squarely in the real of the Foundation legal/privacy team. I will forward your concerns to them. Cheers, Jake

I think the notion of "reach" is linked to our notion of "gap". We see a content "gap" and then we think maybe there is a contributor "gap" underlying it. But maybe there is also a source "gap" which underlies it and so on.

I think the same is true about "reach". While reach is inherently about readers (people), I imagine that (apart from inability to access the web for whatever reason), any shortfall in our reach must relate to people's lack of interest in our content. Given that search engines tend to drive people toward Wikipedia, I don't think lack of awareness of Wikipedia is the major issue here. Lack of interest in our content might relate to a content gap, e.g. we don't have articles about topics they are interested in, or not have it in their language, or because they judge the content to be inferior (or have been pre-conditioned to think it will be inferior, either by personal past experience or teacher/librarian instruction to "not use Wikipedia").

I think we can increase our reach by ensuring that the content we present is as broad and as deep and as authoritative as we can. And obviously more the citations we can provide to authoritative sources and/or online sources the better.

Aside, we do need to convince the librarians about Wikipedia being not as unsafe as they fear. But teachers are a different issue. As a former university professor, the concern about Wikipedia is not necessarily about the quality of the content but that the purpose of the assigned task is to develop the student's own skills in researching and writing (which is not achieved if they just copy from Wikipedia or any other source). As someone who does outreach with university librarians, it is clear that these librarians' position on Wikipedia is now shifting to a more reasonable position of "it's a good place to start but you need to dig deeper using its citations and other sources". Many of them explicitly discuss Wikipedia now as part of workshops they run for students on "how to research an assignment" etc.