This past week in HIST 428, we have been reading about tools and methods for digital history, such as topic modeling and text mining. Our scrapbooks group has also made progress toward completing our Project Contract, and we recently met with Jeanette and Florence Rowe at Agora Coffee downtown to discuss their family and our project.
Two of the most interesting tools and methods that I read about this past week were text mining and topic modeling. Text mining refers to the process of automatically extracting data from a large sample of text in order to work with new information. A great example of text mining is Robert K. Nelson’s “Mining the Dispatch” site,which generates graphs and charts that reveal changing patterns in Civil War newspapers. 
What I found particularly useful about “Mining the Dispatch” is that it allows historians to see what topics were and were of importance to individuals who lived at a certain time period. While these conclusions can always be made through research, text mining helps remove a a potential barrier of modern bias or assumptions about what topics or themes were relevant during a time period and why.
Topic modeling can be thought of what you do with the text you’ve just “mined,” or extracted. According to Megan Brett, topic modeling is “a way to find and trace clusters of words.”  To model a topic, you need:
- A Corpus: A corpus is a large collection of text. (Or, a large body of text.) Just as you need a large sample size to get accurate data when collecting statistics, you need a large body of text to model when attempting to draw connections or parallels.
- Familiarity with the Corpus: While the idea of topic modeling is that a program will help you find themes within a text, the user working with a modeled topic should be familiar with the material themselves.
- A Tool to “Do” The Topic Modeling: To topic model for yourself, Brett suggests adding the MALLET extension to WordPress. This extension does however, require some comfort with Python.
- A Way to Understand Your Results: Once you’ve completed your topic modeling, you will need a way to understand your output. A great way to understand data is to visualize it. One tool that we read about for visualizing data through topic modeling is Voyant.
Voyant is a web-based tool that can be used to visualize data. It is very simple to use. You can either add the URL of a site that you wish to model, or you can paste in a body of text. 
I decided to test out Voyant by entering in my own domain. Voyant gives you a variety of “connections” about your data, but one more visually appealing feature that it has is creating a word cloud of your text. Like other word cloud tools, the words that appear the most are bigger than those that appear more infrequently. Using a slider, you can adjust how many words appear in your word cloud. Unsurprisingly, my word cloud below features words that relate to my coursework and projects.
While I am unsure if our group will use topic modeling in our project, it could be interesting to create a visualization of text based on commonly occurring themes or words from artifacts within the scrapbooks.
Scrapbook Group Update: Meeting the Sisters
On Wednesday, February 5, Emily, Mady, and I met Jeanette Cadwallender and Florence Barnick at Agora Coffee downtown to discuss their family and our project.
We had a great time meeting the sisters, and gained valuable insight into their family and relationship with Fredericksburg. We also learned of valuable organizations and connections that may help us in furthering our research. From our discussion, we learned more about aspects of their family history that cannot be found in the scrapbooks, such as how Anne Rowe Wilson, the sisters mother, met her husband.
Scrapbook Group Update: Social Media
As our group has been working on finalizing our project contract, we decided to make both a project email account and project Instagram. These accounts will help us broaden our research “network” and connect with potential audiences. All group members have access to these accounts and will share management duties.
Our email account, which is email@example.com will allow us to communicate, as necessary, with individuals and organizations related to the Rowe family, as well as with individuals who may be interested in learning about the family.
The Instagram account, which is @rowefamilyscrapbooks will allow us to post updates about artifacts we have found, family members we are learning about, and progress we have made on our project.
Scrapbook Group Update: Understanding Historical Scrapbooks
One point that Jeanette and Florence made to us during our meeting is that they both had no recollection of watching their mother create scrapbooks as individuals might to today. They noted that the artifacts in the scrapbooks were not so much artifacts meant to be saved as part of a narrative or craft project, but rather, simply artifacts that they chose not to get rid of.
Because of this, our group will need to consider the context of “scrapbook keeping” in the nineteenth and twentieth centuries, especially as it compares to modern scrapbook keeping practices. This will not only help our group members understand what we are working with, but also, will provide a framework to those viewing our future website.
 Nelson, Robert K., “Mining the Dispatch,” The University of Richmond Digital Scholarship Lab, accessed February 3, 2020. http://dsl.richmond.edu/dispatch/pages/home
 Brett, Megan R., “Topic Modeling: A Basic Introduction,” Journal of Digital Humanities 2, no. 1 (Winter 2012), accessed February 3, 2020. http://journalofdigitalhumanities.org/2-1/topic-modeling-a-basic-introduction-by-megan-r-brett/
 Stefan Sinclair and Geoffrey Rockwell, “Voyant,” Voyant Tools, 2020, accessed February 3, 2020. https://voyant-tools.org/