the source utilizes a t-SNE algorithm display method which plots readers as x-y coordinates allowing for easy viewer visualization of relationship links using Euclidean distances. The algorithm takes the totals of someone’s reading reviews and relates them to 11 genres, makes their #1 genre their icon and groups them by totals in relation to other genres. The more someone reads of each genre the closer they are to it. The tool of t-SNE makes 11 dimensional data (for the 11 genres) and smooshes it down into 2 dimensional which allows for easier human visualization for the relationships between the genres rather than a one dimensional graph, just a report of the most read genres, or 2 dimensional plots of only 2 genres by their popularity for readers. this allows for many different genre’s relations to be seen all at once and their interconnected relationship to one another
The goal is to see relationships between genre readers in relation to other genres visually. This poses questions in a viewer for the specific relations between different genres and why. Allowing the viewer to see these relationships visually opens for curiosity as to why certain distances and distributions exist (i.e. the distance between romance, mystery, and literary fiction). Perhaps there is a larger societal factor influencing a large amount of space between 2 specific genres(romance and literary fiction), or how certain singular genres can be divided on the other genres a reader tends to drift toward (fantasy).
I thought something that was interesting with the website that we didn’t really touch on in class was the “Readers Mapped by Network” page. The page shows a similar distribution to the “Readers Mapped by Genre-Mix”, however instead of by genre it is by actual book. This allows the viewer to see the relationship between different readers and their most read genres to one another by the actual books they are reading. It connects readers on a 2d plane with Euclidean distances and allows a slide view form the original graph to show the relationship between the two. I think this allows for some great expansion and further thinking on the questions posed by the original graph. However, this expansional graph allows for more data and points to be taken into account with the real connection readers have to one another with the specific book titles. Particularly interesting I think is the greater intermingling of mystery with historical fiction, fantasy, and partially romance. Just looking at the original graph it looks like mystery readers are all out alone, when in reality they are one of the most middle and diverse readers by title to other readers other than fantasy readers. It extracts the skew of the sheer amount of mystery novels some read and allow expansion on the question of how different genres are interconnected and why. It also enforces arguments like we talked about in class with the stigma against romance novels and for literary fiction posed as opposites as on this graph as well they are at distinctly different ends of the graph here as well.
