A Reconsideration of Edges

Posted by on Nov 4, 2013 in Blog, Modeling/Prototyping, Projects

[Submitted by Marc Muschler]

Edges are the connections that can be made within the NewRadial environment which visually demonstrate a relationship between nodes and allow for users to comment on that relation. They are a fundamental aspect of NewRadial through their generation of discourse between users, but this is complicated by the utilization of critical scholarship as nodes within this visual model. Theoretically, the ability to visually indicate the marked connections between critical texts is a promising prospect; one could potentially construct an argument through the use of nodes and edges, writing an essay through the cartography of scholarship that NewRadial can accommodate. That being said, the task of accurately representing these connections as an organized, succinct visual framework is complicated by a number of factors.

Merely determining the direction of the connection is a complicated process that requires a method of indication which is both aesthetically viable and visually accurate. Edges are currently designed as units of connection with no discernible direction. They simply establish a connection between nodes without the use of tools such as arrows to indicate the supposed direction of the argument. The program is designed with the idea in mind that the relationship between source and scholarship is not unidirectional, but bidirectional. Scholarship has as much an effect on the reader as the source text itself, therefore the relationship should not be indicated by arrows but by lines which relate the connection to the user and allow for discussion on that particular edge. However, this concept has come under scrutiny in recent weeks due to the differing relationship between two scholarly texts. One way to consider this predicament is to look at the relationship between two people. For instance, one can say that Charles is the uncle of Sally, but you cannot say that Sally is the uncle of Charles. The connection established does not automatically apply in both directions. So too, does the commentary on a text not affect the subject text in the same as the object text.

A major question that this idea raises is that of affect and the nature of scholarship. Is the relationship between a source text and that of scholarship bi-directional as previously argued, or is this a flawed way of visualizing discourse? What is the affect of scholarship? Does it differ depending on the scholarship, and, if so, how do we distinguish this difference in a visual medium such as NewRadial?  While this is conceptually intriguing, on a practical level one of the possibilities to clarifying this dilemma is to give the user the ability to use uni- or bi-directional edges. This would allow the user to guide the discussion of the relationship between texts without having to navigate the theoretical complications that are implicit in such a discussion. In addition to this, the ability to customize the direction of a relationship between two texts would make NewRadial a more accessible application, allowing for a streamlined, visually comprehendible method of organization that solves some of the issues regarding the visualization of material on a smaller picture level. In this way, up to three separate discussion regarding the relationship between texts can be visualized and accommodated within the NewRadial model, allowing users to engage in extensive and detailed discourse in a dynamic environment.

NewRadial has the potential of bridging the gap of data visualization by operating as a midway point between the massive data clusters representing tens of thousands of nodes of information and the e-reader reading environment where you are interacting with a singular text. Edges are such an integral aspect of the program to consider because the various methods of data visualization infer a different usage for edges within these contexts.  On a big picture level, this conflict raises questions regarding NewRadial’s optimal use: can the program be used to visually represent dozens of texts at once, or would it be more useful as a tool of comparison and contrast between smaller groups of text? Additionally, how can temporality be represented through the use of edges in the NewRadial environment? One can argue that temporality is a necessary factor in the discussion of scholarship and source, but in the context of a data visualization application this problem deserves recognition. Edges will have to adapt to these environments to allow for both accurate visual representation as well as comprehendible connections that promote user discourse regarding the relationships between texts.

1 Comment

  1. Ian Brunton
    November 29, 2013

    One issue to consider in giving users the option to use uni- or bidirectional edges is *how* we present that choice. With bidirectional edges (i.e., as edges currently work in NewRadial), the user simply clicks and drags between two nodes. Easy as pie. With unidirectional edges, however, the user will have to specify which node is the object and which is the subject. The simplest way to do this is to designate the first node clicked on as the subject and the second as the object. But this requires the user to know that that’s what will happen; what if she changes her mind, or realises she clicked in the wrong order? On the other hand, if the choice is presented to the user after both nodes have been selected, we will have to provide some mechanism in the UI to (a) choose between a uni- and bidirectional edge, and (b) if the former, specify the relationship between the two, i.e., which one is Charles and which one is Sally. Part of the challenge will be providing sufficient choice and information, without letting the UI become cluttered and bewildering.


Leave a Reply