by Silvio Lorusso and Ania Molenda
Within the Platforms group three independent online publishers: Amateur Cities, Online Open! and Open Set, have been developing a prototype of an algorithmic tool allowing small online publishing platforms to federate and expand readership. The tool, which we called Platform 2 Platform, establishes relations between the content (including articles, images, videos and the like) by crawling each platform’s website. Those relations are based on text analysis and a set of criteria defined by the editors such as tags and bibliography. Before recommending federated content to their readers, the publishers are able to rate the relevance and quality of an automated suggestion. In this way the tool feeds into the editorial process instead of automatically suggesting content based on readers’ individual preferences and browsing history, such as in the case of clickbait. By giving the editors’ agency in the recommendation process, it stands against the homogenization of content presented to users, which often leads to biases and filter bubbles. The tool is designed to strengthen the position of the editor, not to replace it with an automated solution. It gives editors more power in suggesting meaningful content to the readers and providing context to the content they publish, both inside and outside of their own platform. While engaging in a collective project, individual platforms can still remain in control of their vision, content and the links to other platforms that are created. It is an important aspect that Platform 2 Platform can stimulate and create relationships not only between content but also between people engaged in publishing, and strengthen the community of independent publishing platforms.
An Open Relation
With Platform 2 Platform we want to advocate for an open relation between the editors, algorithm(s) and the readers. We want to open up the recommendation black box and make its logic both visible and legible. Publishers are able to accept or reject suggestions, based not only on relevance, but also on expected quality standards and adherence to the curatorial and artistic vision of the platform. They can also assess the quality and relevance of the automated suggestions in order to improve them in the long run. We value the general understanding of the algorithm’s recommendations, that’s why the latter has been conceptualized by all team members including the publishers, a designer and a developer. Eventually relations created by the tool and the reasoning behind them are presented alongside the original platforms’ content, promoting transparency of the rules based on which the link has been established. So, both editorial decisions and the algorithmic logic become visible to the reader. The legibility of the algorithm provides a novel look into the content, and fosters reflexivity on different types of meaningful relations that can be created between platforms and what they publish.
Small independent publishers often rely on corporate solutions including social media and search engines in order to disseminate their content. They can choose to use sponsored posts or concentrate on improving Search Engine Optimization (SEO) to jump the rank of the search query. In both cases, publishers have to deal with a high degree of uncertainty when it comes to the way, in which their content is disseminated and positioned. Platform 2 Platform aims to nurture what can be called a counternetwork – a network in which each and every node has an agency. While the success of dissemination of content through social media and search engines is determined by economic power (e.g. the ability to finance sponsored posts) and IT skills (to outsmart the PageRank algorithm), Platform 2 Platform produces a semi-egalitarian ecosystem (where each article is accompanied by a series of suggestions) based on trust, where each platform that joins knows that it will host suggestions to other platforms’ content. In this way platforms can support one another in disseminating content, reaching a wider audience, and diminish their reliance on corporate solutions. Additionally by creating links between platforms and driving click rates they can help improve their contents’ ranking. – We see it as a form of positive reinterpretation of clickbait.
Since the platforms participating in the Making Public project focus on knowledge creation and dissemination rather than making news they are likely to remain on the margins of popularity. They do not publish shocking or viral content that is meant to attract views in terms of numbers of visits and clicks, but to create insights and stimulate thoughtful reflection. This issue became the underlying logic for the tool’s attempt to break through the scheme that promotes shocking content, and amplify the reach of small platforms without pushing them to compromise the quality of the content. Instead of promoting content that is more interesting commercially, Platform 2 Platform uses the aspect of federation and trust to support and popularize independent publishing.
Levels of Trust
Platform 2 Platform is based on a social notion of trust. By joining the federation, each platform commits to a collective project where trust in the other platforms’ values and attitude are central to the success of the common endeavor. Even though every platform has full editorial and curatorial control, it needs to be open to allow content that has been created according to different editorial standards or presents a different point of view. In that sense, by trusting in the quality of content and editorial work of other platforms, the members of the independent publishers’ network create a sense of credibility towards and for one another.
P2P and Federation
Last but not least, Platform 2 Platform brings together aspects of peer-2-peer (P2P) philosophy and platform federation. The notion which it borrows from P2P is that each platform remains in direct dialogue with one another, by reviewing and revealing the suggestions. It increases the knowledge that platforms have about the work of others and stimulates collaboration. From platform federation, it borrows the fact that platforms maintain self-governance. They can individually decide which links to include and which not. While incorporating external content, each platform maintains its own identity and curatorial or editorial strategy.
Right now, there are three publishing platforms involved in the project. At this stage all of them have been able to meet in person and discuss in detail what their objectives and needs are, and translate that to the functioning of the tool. What would it mean to open this counternetwork to more platforms? What kind of method of determining consensus would have be to applied to accommodate the needs of new members? While it is relatively easy to discuss specific needs between three platforms, it becomes increasingly difficult with a bigger group. How to deal with the aspect of scale, is a question that has not been addressed at this stage of the project, but remains an important consideration for its future development.