Spring 2026 Spatial Representation & Analysis of Archives
Richard
Spring 2026: Showcase – Spatial Representation & Analysis of Archives
On May 15, 2026, graduate students in the INS608D class (“Spatial Representation & Analysis for Library & Archive Collections”) at the University of Maryland showcased original and innovative work.
This course was a practical exploration of geospatial techniques to manipulate and represent library and archival collections. Techniques include digitization of maps, geocoding, geolocating, georeferencing, vectorization, spatial analysis, 3D maps, LLMs and spatial data, computational analysis, and representation of library and archival content using open source and commercial tools, towards developing new access interfaces.
This course is part of a new series of 5-courses in Computational Archival Science (CAS) (see: https://ai-collaboratory.net/cas/). These courses are specifically designed for MLIS students interested in developing skills in digital curation and computational thinking but are also suited to INFO master’s students in other programs and doctoral students, as well as graduate students from other colleges. They currently include:
Coding for Non-Coders: LBSC708F: Introduction to Computational Archival Science (CAS) & Python
GenAI & LLMs: INST728L: GenAI & Large Language Models (LLMs) for Library and Archive Collections
Digital Curation/Data Science: INST742: Implementing Digital Curation
Graphs: INST608D: Using Network Visualization to Explore Library & Archive Collections
Students applied geospatial techniques throughout the semester using the Asheville, NC Urban Renewal Project’s archival records. This is part of an AIC project called Measuring the Impact of Urban Renewal. Map samples from these experiments include:
A three-week final project allowed students to showcase their own projects:
Kurosh JAFARI: StoryMap for Geolocating Historical Manuscripts from the Library of Congress
Bill STEA: A Map of Mythical, Paranormal, and Supernatural Folklore in Southern Maryland
Ana STARR: Mapping and Analyzing Asheville’s Owners in Polygons
Ben POLLOCK: An Exploration of Beach Haven’s Historic District
Noah JONES: The “Roaming” Emperor: Mapping the Travels of the Roman Emperor Hadrian
Samuel BESSEN: Mapping Baltimore’s Historic Sheet Music Publishers
Melanie BELKIN: Let’s Go Back to Rockville: A Community-Based QGIS Exploration of Montgomery County, Maryland
Shameem RAZACK: Embroidering Transit Density: Mapping Public Transportation and Material Data Visualization in Cook County
1. StoryMap for Geolocating Historical Manuscripts from the Library of Congress
Author: Kurosh JAFARI
Abstract: I am working on a story map for the Persian manuscripts housed at the Library of Congress. There are roughly 100 manuscripts in the library’s Persian Language Rare Materials online digital collection. My story map will discuss the history of the collection at the library, the geographic scope of where the manuscripts were written, and collection highlights in different subject areas.
Dataset: I had to create my own data set on a spreadsheet for this project as none did not exist. I included the title, author and scribe (if known), URL, notes on the manuscript, where it was written, latitude and longitude geolocated/geocoded coordinates. All of the information will come from the Persian Language Rare Materials online digital collection. I may use historical maps? Since this is also an internship project, I can only use Library of Congress materials to do this project.
Tools: I will be using ArcGIS for this project as this is what the Library of Congress uses to create story maps. QGIS is only used by GIS librarians and I could not get permission to use it at the library. I will also be using the Library of Congress’ online digital collections for the metadata information for this project and included images in the story map.
Video: TBPosted
Video: A Storymap on the Persian manuscripts in the libraries collection and a map showing the geographical distribution of where they were written. The map will be generated via a csv file of the spreadsheet I created (mentioned above). I will then record a video discussing my story map.
2. A Map of Mythical, Paranormal, and Supernatural Folklore in Southern Maryland
Author: Bill STEA
Abstract: I would like to create a map based on a unique dataset: locations of mythical and supernatural folklore in St. Mary’s County, MD. I will use QGIS’s Open Street Map and then create vectors that outline specific areas in which there have been reports of hauntings, strange creatures, and other unusual occurrences. The dataset in the attributes table would involve specific pinpoints such as buildings and landmarks (ex. The Point Lookout Lighthouse, Moll Dyer Rock, etc.) as well as some areas that may have a wider area (ex. Sotterly Hill). Each area would also contain some metadata summarizing the unique story of each point of reference. I will also create a raster layer of an older map of Point Lookout’s Civil War camp and place it over the QGIS Open Street layer. The map would create context for some of the local culture that has grown in these counties as these stories continue to be retold and passed down from generation to generation.
Sources & Datasets: (1) Monsters of Maryland : mysterious creatures in the Old Line State by Ed Okonowicz; (2) Haunted Southern Maryland (Haunted America), by David Thompson; (3) Ghosthunting Maryland, by Michael J. Varhola; (4) A map of Camp Lincoln in Point Lookout.
Tools & Approach: I will begin with a QGIS open street map of the St. Mary’s County area, and then create a layer that will add pins to specific locations. I will create a polygon around certain structures and buildings where appropriate, and will create larger polygons that may overlap for wider areas (forests, rivers, etc.). Once that is complete, I will attempt to move this over to Google Earth and create a video tour of some of the areas.
Deliverables: (1) A set of locations and relevant metadata found during the research; (2) A screenshot of Point Lookout, along with an overlaid raster layer of the older map; (3) A video presentation to examine the results through QGIS.
3. Mapping and Analyzing Asheville’s Owners in Polygons
Author: Ana STARR
Abstract: I would like to continue working with the Asheville data as I have for the last few semesters. Building on some of the layers we explored in class, I would like to use some of the Owners data to map and label owners.This would build on last semester’s network class and give me another visual of where some of those names/parcels I grew familiar with are actually located, and how they’ve changed over time. I want to also explore more of the spatial analysis techniques in the Toolbox we used for assignment 6. To limit the scope I thought I’d stick to the area within one of the polygons we created earlier in the semester 56,57,58.
Dataset: Owners data, UR Parcel layer, 1966 land acquisition layer, 1950 census map layer, current parcels layer.
Tools: QGIS, latlong.net
Deliverables: QGIS layers/images, revised owners data set, QGIS Screenshots, toolbox analysis outcomes, Video walkthrough of mapping and analysis.
Video: TBPosted
4. An Exploration of Beach Haven’s Historic District
Author: Ben POLLOCK
Abstract: I have chosen to apply the skills I learned in this course to highlight the historic district of Beach Haven, NJ. To do this, I will use the OpenStreet layer in QGIS, then apply a Sanborn fire map of the district from 1911. I will be using georeferencing, geolocating, and vectorization. I also plan to utilize TerraLab, the AI platform that we discussed for Assignment 8. With these components altogether, I would like to have the start of a map where individual properties can be selected to reveal a brief history of the property and a comparison of the property in 1911 to 2026.
Dataset: QGIS, 1911 Beach Haven Sanborn Map (Princeton University Library), National Register of Historic Places Registration Form (Internet Archive), Other archival resources (Long Beach Island Historical Museum)
Tools: (1) QGIS Layers – Application of OpenStreet map, Sanborn map, (2) Georeferencing – Snapping and georeferencing Sanborn map to Beach Haven Historic District, (3) Terralab – Segmentation of land parcels and individual properties
Deliverables: (1) A presentation with a proposal to utilize the skills, datasets, and tools mentioned above for a longer-term project. (2) A QGIS map the georeferenced 1911 Sanborn map, one segmented parcel, and at least one segmented property. (3) Possible Addition: At least one segmented property with geolocation and brief history.
Video: TBPosted
5. The “Roaming” Emperor: Mapping the Travels of the Roman Emperor Hadrian
Author: Noah JONES
Abstract: I have always been obsessed with the classical world, so for my final assignment, I would like to focus on the travels of the 2nd-century Roman emperor Hadrian. Ruling at the peak of the Roman Empire’s power and prosperity, Hadrian is famous for visiting nearly every province of the empire during his over 20-year-long reign. I intend to plot out the cities he visited in Europe, Africa, and Asia throughout his reign as a means of both highlighting the wide breadth of the Rome’s domain at it’s height and of giving a sense of what cities were prominent during this period of Mediterranean history.
Dataset: Carole Raddato, a classics scholar and Hadrian fanatic who has actually made an effort to visit almost all of the places the emperor did herself, has created a detailed list of the cities he visited on her website, Following Hadrian. I will also make use of various layer files posted on Github that include a map of the Roman Empire, as well as its provinces and cities.
Tools: I installed the plugin Quick Map Service in QGIS in order to get a clean map with no modern day borders or country names. On top of that, I will add layers depicting a map of the empire during Hadrian’s reign, labels for the provinces, and labels for the cities in order to create an era-accurate grounding for the data. From there, I intend to plot the cities he visited.
Deliverables: I will take multiple screenshots of my work in QGIS and make a video going over Hadrian’s travels using it.
Abstract: During the late 19th and early 20th century, glass manufacturing rapidly expanded in the United States, and became more concentrated in states like West Virginia and Indiana due to the energy resources in those regions and the expansion of railroads to make transportation of materials easier. I will track this growth and the simultaneous growth of supply lines.
Dataset: I am collecting data from the 1880 Census Report on Glass Manufacturing and the 1919 Directory of Glass Factories in the US. I am using Jeremy Atack’s timestamped railroad shp file.
Tools: I have collected county-level data about glass manufacturers for the years 1880 and 1919. I will create color-coded layers displaying this data in various ways, using color to show both the number of manufacturers, the number of glass furnaces, and growth over time for particular ranges. I will also overlay a railroad map timestamped to 1880 and 1919 to show how the growth of railroads impacted the expansion of glass manufacturing.
Deliverables: Metadata and statistics about glass production in 1880 and 1919. Screenshots of various ways of displaying statistical data about glass production overlayed with railroad lines. A video presentation using the map to show the development of glass production, with a discussion about other ways that data could be visualized to flesh out the history of glass.
Video: TB Posted
7. Mapping Baltimore’s Historic Sheet Music Publishers
Authors: Samuel BESSEN
Abstract: My final project for this course will be to apply the GIS skills I’ve learned to a new dataset: historic sheet music publishers in Baltimore City. From the mid-18th century into the 19th century, Baltimore was a hub of popular American sheet music production, along with New York, Philadelphia, and Boston. In 2021, I began a project called Mapping Tin Pan Alley, using data about New York publishers to visualize the industry from 1880-1940. This visualization proved useful in tracking the migration of the industry Northward through the city. This project would apply a similar lens to Baltimore, examining these publishers in the context of the city in which they thrived.
Dataset: Data about these publishers is spread over a variety of primary and secondary sources. I will use the following sources to determine publisher addresses and create a dataset. Primary Sources: (1) Historic sheet music, which often contains publisher addresses, (2) Historic Baltimore City directories. Secondary Sources: (1) Dichter, H., Wright, E. A., McDevitt, J. A., & Shapiro, E. (1941). Early American sheet music : its lure and its lore, 1768-1889. R.R. Bowker Co. (2) Sonneck, O. G., & Lowens, I. (1964). A bibliography of early secular American music, 18th century(W. T. Upton, Ed.; Revised and enlarged). Da Capo Press.
Tools: Using QGIS, I will start with a default street map. Above that, I will geo-reference the 1832 Fielding map of Baltimore City, which best matches the time period of the dataset. This will show any changes to street names or locations in the city. I will then geo-locate the above addresses using LatLong.net and Google Maps, adding a point layer above the historic map.
Deliverables: (1) Original dataset of historic publisher addresses, (2) 2-3 screenshots of QGIS project with three layers: (a) Open Street Map, (b) Geo-referenced 1832 Fielding map, (c) Points layer of labeled sheet music publishers, (d) Video demonstration of results, including a short discussion of possible future developments for the project
8. Let’s Go Back to Rockville: A Community-Based QGIS Exploration of Montgomery County, Maryland
Author: Melanie BELKIN
Abstract: My project will be a largely exploratory experiment to map out the community resources of my hometown and potentially look at the historical development of the county as a whole. I’ve found number of datasets of points-based information, including schools, veterinary hospitals, hospitals, libraries, and farmers markets in the area.I’m also interested in mapping historical documents onto the modern terrain of rockville. I’m worried most of the datasets I found are point-based and don’t lend themselves to the polygon techniques we used for Asheville parcels, but perhaps that is not necessary for this final. I don’t expect to have any groundbreaking findings, I just think it would be neat to explore an area I’m familiar with while utilizing the skills I’ve developed during this course. I was actually interested in viewing a timeline of the neighborhood of my childhood home developing over time, but I was unable to find a good and consistent source for specific neighborhoods.
Tools: QGIS, using the toolsets demonstrated throughout the course – point data sets with conditional displays, projecting png images onto the streetmap layer
Deliverables: QGIS layer screenshots
Video: TBPosted
9. Embroidering Transit Density: Mapping Public Transportation and Material Data Visualization in Cook County
Author: Shameem RAZACK
Abstract: For my final project for this course, I will expand on the research and visualizations developed in my paper on public transportation commuters in Cicero Township and Evanston Township, Cook County, Illinois. Using data from the 2023 U.S. Census Bureau’s American Community Survey (ACS) Table S0802, specifically the “Public transportation (excluding taxicab)” category, I will use QGIS to spatially visualize commuter density alongside the embroidered visualizations already produced for the project. While the original paper explored how stacked bar charts and embroidery could function as counter-visualizations that challenge the neutrality of statistical representation, this project introduces a geographic and spatial dimension through GIS mapping. By mapping transportation data onto CTA transit infrastructure in Cook County, I aim to examine how commuter density, race, class, and urban mobility are spatially organized and materially experienced across different suburban geographies. The project will continue to draw from Black studies, critical data studies, and Black geographies approaches to visualization, particularly the critiques of classification, enumeration, and “coded inequity” raised by Ruha Benjamin, Mimi Onuoha, and others.
Cicero Township public transportation commuters: 3,682 commuters
Evanston Township public transportation commuters: 5,608 commuters
Additional spatial data will include:
CTA rail line shapefiles (Pink Line and Purple Line)
Cook County municipal boundary shapefiles
OpenStreetMap base layers
Geospatial reference points corresponding to Cicero and Evanston transit corridors
Tools: Using QGIS, I will create layered spatial visualizations that connect commuter density to transit infrastructures and suburban geography. I will begin with an OpenStreetMap base layer and add municipal boundaries for Cicero Township and Evanston Township. Above these layers, I will incorporate CTA rail line data for the Pink and Purple Line routes referenced in the embroidered visualizations.I will then create proportional symbol layers or density visualizations representing commuter counts derived from ACS data. These spatial visualizations will be placed alongside photographs of the embroidered fishbone stitch visualizations already produced in the paper. The embroidery component translates commuter density into tactile, layered material form, with one stitch equaling 40 commuters. The goal is to place statistical mapping into conversation with material and embodied forms of knowledge production, showing how transit density can be understood both spatially and tactually through mapping and textile visualization.Deliverable:
QGIS project file with layered spatial visualizations
2–3 screenshots of completed QGIS maps featuring:
OpenStreetMap base layer
Cook County township boundaries
CTA transit line layers
Public transportation commuter density visualizations
Photographic documentation of embroidered visualizations from the paper
Short video demonstration explaining the QGIS layers and visualization process.
