Wikimedia Product/Perspectives/Culture/Content Gaps

Content Gaps

Summary

Are we there yet? Making the sum of the world’s knowledge accessible to all the world’s people presumes that the extent of knowledge can be described,^[1] measured,^[2] and tracked.^[3] It also implies that content gaps can be identified and filled with context-relevant takes on the missing knowledge – this may be the most difficult part of the problem. For our purposes, gap knowledge can be said to fall into one of these general categories: potentially portable knowledge,^[4] contextually nuanced knowledge^[5] or contextually anchored knowledge.^[6] The three types require different intervention strategies, and may have different relative priority for a given a regional community. For this reason, intervention strategies (product features, global tools, community mobilization) must be developed in parallel. Solutions for filling content gaps also surface new challenges around notability and verifiability^[7] – how must current standards for authority need to evolve in order to encourage the growth of each type of gap knowledge? And finally, assuming that it is possible to close the content gap through various means, what support mechanisms will be required to maintain this body of knowledge? It will be necessary to have tools that monitor and signal the freshness of content to the community best suited to maintain it. In sum, making all knowledge available to all the world’s people demands the organization and participation of all the world’s peoples, and interventions deployed at pace with the rate of growth of the overall pool of world knowledge. [3]

White Paper

DRAFT

Resources

L. Zia and N. Schaaf, 2015 Gapfinder https://www.mediawiki.org/wiki/GapFinder

L. Zia, 2017 Article Expansion Recommendation System https://www.mediawiki.org/wiki/Wikimedia_Research/Showcase#December_2017

A. Halfaker R.S. Geiger, J. Morgan and J. Reidl, 2018 The Rise and Decline of an Open Collaboration System: How Wikipedia’s reaction to popularity is causing its decline https://drive.google.com/open?id=1FzMCptTMOvvwRRGQPyh2AsBrGneOIG3F

2017 Movement Strategy Report: New Voices Synthesis https://meta.wikimedia.org/wiki/Strategy/Wikimedia_movement/2017/Sources/New_Voices_Synthesis_report_(July_2017)#How_to_attract_new_users

References

↑ T. Negrin: “The Static Knowledge Pie”: using the sum of all topics currently covered on all wikis as a measure of completeness, the content gap is measured as topics existing on some projects and absent on others. Requires an exhaustive but extensible topic model.
↑ T. Negrin: “The Potential Knowledge Pie”: The Static Knowledge Pie with the addition of known gaps across all wikis (e.g.estimated number of articles there should be about female scientists).
↑ T. Negrin: “The Dynamic Knowledge Pie”: the diameter of the pie increases over time making the content gap a dynamic function of the relationship between rate of knowledge created in the world and knowledge captured on the system.
↑ Potentially Portability reflects the degree to which certain classes of knowledge have the potential to be “ported” or translated with low loss in fidelity. Scientific terminology and topics are an example of PPK. Certain types of legacy media (e.g. illuminated manuscripts) that can be ported from one medium to another could also be considered PPK. Transferring Potentially Portable topics and formats is challenging, but methods for tackling them are known unknowns.
↑ Contextually Nuanced knowledge is a category of knowledge that requires a higher degree of human input to transfer (e.g. subject matter expertise, topical knowledge and interpretation). The history of the Korean War written from a Japanese, or American or Korean point of view would be an example of CNK.
↑ Contextually Anchored knowledge is a category of knowledge that is utilized by a specific community in a specific way. This knowledge can be described in other contexts, but not utilized in a comparable way. The Hawaiian Oli chant is an example of CAK.
↑ Decline of editor retention in mature language communities on Wiki being (partially) caused by lack of welcome for newcomers

[1] T. Negrin: “The Static Knowledge Pie”: using the sum of all topics currently covered on all wikis as a measure of completeness, the content gap is measured as topics existing on some projects and absent on others. Requires an exhaustive but extensible topic model.

[2] T. Negrin: “The Potential Knowledge Pie”: The Static Knowledge Pie with the addition of known gaps across all wikis (e.g.estimated number of articles there should be about female scientists).

[3] T. Negrin: “The Dynamic Knowledge Pie”: the diameter of the pie increases over time making the content gap a dynamic function of the relationship between rate of knowledge created in the world and knowledge captured on the system.

[4] Potentially Portability reflects the degree to which certain classes of knowledge have the potential to be “ported” or translated with low loss in fidelity. Scientific terminology and topics are an example of PPK. Certain types of legacy media (e.g. illuminated manuscripts) that can be ported from one medium to another could also be considered PPK. Transferring Potentially Portable topics and formats is challenging, but methods for tackling them are known unknowns.

[5] Contextually Nuanced knowledge is a category of knowledge that requires a higher degree of human input to transfer (e.g. subject matter expertise, topical knowledge and interpretation). The history of the Korean War written from a Japanese, or American or Korean point of view would be an example of CNK.

[6] Contextually Anchored knowledge is a category of knowledge that is utilized by a specific community in a specific way. This knowledge can be described in other contexts, but not utilized in a comparable way. The Hawaiian Oli chant is an example of CAK.

[7] Decline of editor retention in mature language communities on Wiki being (partially) caused by lack of welcome for newcomers

[1]

[2]

[3]

[4]

[5]

[6]

[7]