This paper details the process of creating a workflow to extract, structure, and categorize the text of historic-period cookbooks for future analysis using visualization applications. This includes a consideration of bias in both data science and data analysis projects. After design of the initial database and workflow, the project went through three additional rounds of testing that showed the possible limitations of the initial schema and controlled vocabularies and led to specific changes in the database. The paper considers how the database can be used to ask questions of the data, and presents possible strategies for answering these research questions. The paper concludes with a brief discussion of the possible future for textually based data science projects and cookbook analysis.