Data Visualizations
Network Graphs of Dickinson Variants
Interactive network graph utilizes nodes (points) and edges (lines) to represent Dickinson textual variants as relational textual entities. The visual representations in this section allow easy zooming and panning to view details or explore different parts of the network. Hovering over or clicking on nodes can display additional information, such as details about the entity or the nature of the relationship. Let us consider some examples to better understand how this works.Base Text vs Variant: "souls"
We begin variant visualization by looking at how a base text connects with a variant or vise versa in a poem or across poems in the collection. We take the poem title "Bereavement in their death to feel" as an example. Here, we have the base text "souls" with three variants marked against it - "World", "selves", and "Sun". The node "souls" is connected to each of its variants in the poem. The nodes are color-coded. The light blue for base text and light green for variants.
Base Text vs Variant: "Sun"
The network graph below visualizes the
connection
involving the word "Sun" in the dataset. It highlights how the word "Sun", a direct variant of "souls", is connected to various base text nodes. "Sun" here has extended connections with other words such as "spell", "sustain", and Noon" within the context of different poems. In two occasions, "Sun" serves as base text for "spell" and "sustain" while it becomes a variant for "Noon".In this graph, the light blue nodes represent base texts while the orange nodes represent variants. The edges (lines) between nodes indicate a connection between the base texts and variants.
Base Text vs Variant: Fascicle 34
Below is a basic network graph comparing the Base Text and variants entries for Fascicle 34.
In this visualization:
- The light blue nodes represent unique words or phrases from the Base Text column.
- The light green nodes represent unique words or phrases from the variant column.
- The edges (lines) between nodes indicate an association between the Base Text and variants entries within the same poem.
Interactive Network Graph by Year
Below is an example of an interactive network graph highlighting base text and variants connection. The complex relationships through nodes and edges represent the unique textual composition of words by Dickinson. The structure of the connection denotes that nodes (base text and variants) that are grouped together are similar in meaning and are strongly related. These help to create clusters or communities within the graph, revealing groups of nodes that are more closely related to each other than to the rest of the network. Within a cluster, some nodes are influential because of their high degree of centrality within a network.
Weighted Network Graph by Year and Fascicle
Weighted network graph helps us to visualize base text and variants representing the strength or frequency of their connections. This is expressed between nodes with varying thickness or colors of the edges. The first step is to extract and count word pairs in the collections. For each text, count how frequently each pair of words co-occurs. Next, add edges to the pairs (nodes) with weights corresponding to these frequencies. The visualizations below vividly represent weighted base-text and variants co-occurrences.
Learn more by visiting Creating Interactive Visualizations with Plotly by Programming Historian