As an undergraduate in the early 1980s, I could organize my research with just handwritten notes and a good memory. After I returned to academia to pursue a graduate degree, I quickly discovered I needed a new approach. Fortunately, I'd already become familiar with the potential of Adobe's Lightroom (see sidebar) for organizing large photo collections, and quickly adapted it to meet my research needs.

My research seminar project included an analysis of 1,600 pages of newspaper articles which I collected from microfilm scans and an online source. My organizational solution drew on my photography background. I have used Lightroom for several years to manage photography projects and have worked with over 75,000 digital images in multiple collections. I knew that I could use Lightroom to add extensive notes about each article in the caption field. Additionally, Lightroom's integrated system of keywords, colors, stars, and flags would allow me pull together different groupings for analysis and quantify trends. Lightroom's advanced search functions would, I knew, allow me to quickly retrieve and associate articles across a range of categories.

Before I collected the scans, I developed a computer folder structure and file naming convention so that I could optimize Lightroom's search features. I also considered what type of searches I might want to perform and built a preliminary keyword list based on my anticipated search needs. Both of these steps would apply to any research project involving multiple digital images. The folder and file name should be carefully designed to help the user quickly find specific images. The keyword list resembles a subject index and helps the user find files with shared attributes, even if they are spread across multiple folders.

The research topics and questions should guide the development of the folder structure. Picking a single guiding principle—be it topical, geographic, or chronological—makes it easier to determine where to place each document. For instance, research on events in China during the reign of the Guangxu Emperor suggests a top level folder system with event names such as the Hundred Days' Reform and the Boxer Uprising. Research on international conflicts in 1900 suggests a top level folder system with country names such as China, the Philippines, and South Africa. Research on events that influenced U.S. immigration quotas of Chinese citizens from the Chinese Exclusion Act of 1882 to the National Origins Act of 1924 suggests a date-based approach. Other ideas for top level folders include source types, authors, government departments, political parties, building designs, or art forms. At this level, we haven't departed too far from a familiar hierarchical file-folder system. But within Lightroom, we can now take steps to creatively expand our thinking about classifications.

Early in the project, decide on a file naming convention that will help you find and group files. I prefer to assign file names that identify the folder location. My research compares a Democratic newspaper to a Republican newspaper during a 1901 city election period and asks how they differed in their reporting on topics associated with German ethnicity. The top folders consists of the papers' names, the next levels include the year and month and then the day. The file name "FWS 1901-03-05-p04" identifies the top folder "The Fort Wayne Sentinel", the subfolder "March 1901", the bottom level folder "March 5th" and the image of page four. Placing the "p" in front of the page number lets me group all the page fours (the editorial page) together with a simple search of "p04" while excluding all the April dates or the 4th of the months.

Keywords (sometimes called "tags") allow you to break out of the hierarchical file structure and group documents in more fluid and useful ways. Lightroom's keywording features include batch or single image application, autocomplete, leveled keywords and the display of the keyword tallies. Typically, I develop a general list of keywords and modify the list after I process a test group. For my newspaper research, I used keywords to categorize the article contents and to rate the articles in a continuum of positive, neutral, apologist and negative. In one section, I worked with ninety articles about Germany. My analysis included a quantitative comparison of the keyword tone ratings that I assigned. With simple searches, I could view the rating distribution for subtopics such as Germany's role in China, Germany's leader, Kaiser Wilhem II, and remarks about Germany on the editorial page by newspaper.

I also make use of the ability to level keywords. Lightroom allows you to create keyword structures in a manner similar to Library of Congress subject headings. For instance, I am working on a large database of student travel images. To identify image location, the keyword structure begins with the continent, then country, state and city. Sophisticated searches can therefore be devised to zero in on a country, region, or particular city, or, alternatively, to retrieve photos from a broad region but exclude certain locales.

So far, I have described three ways of classifying and searching for an image—folder, filename, keyword. Since Lightroom allows you to save searches and create "virtual folders," a researcher can have as many combined classifications as she has ideas. I was analyzing how different newspapers reported on several topics, and keywords helped me separate the articles about each topic from the pages and compare the tone of the articles between Republican and Democratic newspapers. As I read the articles on Germany, I wrote notes in the caption field and assigned my tone rating as a keyword. I also labeled articles that I wanted to quote by assigning them a color—in this case it was green—adding another level of classification and retrievability.

After reading the articles, I had a good idea of how I wanted to break the Germany section into subtopics. I used LR to search my notes for China and Kaiser. Each search brought together the articles and displayed the keyword tone distribution for the subtopic. Adding the newspaper abbreviation in the search line displayed the keyword tone totals by newspaper. The keyword tallies revealed that the Republican paper consistently used a more negative tone than the Democratic paper across all the subtopics regarding Germany. Additionally, although both parties nominated German American candidates for mayor, only the traditionally German American Democratic party tempered the negative national coverage about Germany with apologist statements from the German press.

After I built some simple comparative charts in Excel based on tone rating percentages, I began writing. When I wanted to refer to an article, a simple click displayed the article full size on the screen. When I quoted an article, I added a white "pick" flag so that I could later add the citation information to EndNote as one step.

Lightroom allowed me to intelligently manage over 1,600 images using several different overlapping methods—folder, filename, keyword, color—and I haven't even had space to get into how I used the print features. Without the burden of having to sift through paper notes or notes arranged in a linear word processing document, I could concentrate on analyzing the articles and making concrete, empirical, and quantifiable discoveries. It took some planning and diligence, but the end result was a new window on my topic that wouldn't otherwise have been possible for me, a graduate student with limited time. Without such a tool in my repertoire, I wouldn't have attempted the project in a semester.

Nancy Brown is a history graduate student at Indiana University-Purdue University Indianapolis. She anticipates graduating in 2013.