So, yesterday we left off establishing the Steemit blockchain as a set of many possible graphs and each piece of content being a node within each of these graphs. Each individual graph serving a some sort of relationship between the nodes that we consider. By traversing an edge on one of these graphs we can discover new content from one node to another using some relationship we consider. This idea forms the current basis of Project Greenshift.
Why Greenshift? Well, green represents money and shifting represents moving that money. Putting that altogether, the hope is to distribute rewards more evenly and to more people via giving users new avenues to explore and discover new and exciting content that matches their needs. Rather than seek some generalized equation, we seek something a little more personal.
So, how are we going to span the network and achieve this goal? Initially, we need to select several different relationships between content that allow users some room to explore without being confined to a singular topic or author or monetary value. Right now, I'm leaning towards figuring out five different relationships and using that relationship to find five different but related pieces of content.
Then the user will have the option of selecting one of those pieces of content where they will be able to view the piece of content and then have 5 more pieces of content uncovered using the newly selected piece content as a new reference for the relationship. Of course, we'll need to have some backups and conditional logic to avoid selecting content users have already looked at, but that's something we'll have to experiment with and test out.
The personal aspect of such a means of traversal is that not every user will select the same piece of content out of the five allowing for branching and exploration which is different to scrolling down the linear paths established on current interfaces. Essentially, a super basic recommendation engine which later down the road we can add different algorithms to personalize traversal across the Steemit blockchain in a less linear manner.
Another benefit to exploring the network in such a manner as opposed to filters in that minimal user input and effort will be required to explore the blockchain. The problem with a lot of filters is that they cut across the space linearly using a single parameter.
With the proposed method of travel, such criteria is irrelevant, only that some relationship exists with the current piece of content being evaluated. Such an approach is localized versus the global approach of current filters in existence. And a localized approach would appear to provide a easier means to create a personalized approach to content discovery than an overarching global filter or rating system.
So, we've explored this idea in a little more detail and are starting to flesh some of the more basic implementation details. Next time, we will explore some different relationships and look at them individuality and in more detail.
I can't promise that this will be a daily blog, but I want to try to dedicate a little time each day to exploring this idea and writing and documenting this process helps me to brainstorm the idea and perhaps if anyone is reading feedback can be added. Although I'll still be happy to write this to nobody. It means I don't have to proofread!
