Archive for the ‘Linked Data’ Category

  • Link to Data Dump/Analyze/Display Session


    Hi ThatCampers,

    Here is the link to the document where we gathered recommended programs to approach this problem with modular solutions.

    At this point there is not one overall program that we could come up with to do all these things at once.

    I am interested as a person squarely in the humanities to partner with an IT/library science people to help to develop something that would be usable and very user friendly for humanities-types.





  • Connecting the personal story to public history for academia and journalism


    I have been interested in connecting an individual, personal story to a public sense of histories of places, events, timelines. That is, I think there is a potential paradigm for threading stories, histories, archives and communities.

    There is nothing new about this thought. Fields like oral histories/public history have well established research around these themes. But I want to put this proposal out there as a part-talk, part-make, part-teach session. I have very, very tiny seeds of the project that I can share, and it will be good to brainstorm around it, or see if there are folks interested in collaborating, too.

    I am very much an outsider to digital humanities but I am hoping to learn and share (I am a research engineer who works with x-ray optics, and a journalist/amateur oral historian). I had been collecting oral histories at a physics research lab for half of 2009. Felt like it told a rich layered history of state-individual collaborations in science. That, then, broadened to interest in the intersection of scientific research and people’s histories….which then broadened to oral histories in other fields of work. The seeds I can talk about or illustrate are histories from two physicists who discuss the origin of a set of champagne bottles in a particle accelerator control room, and some other stories on dying traditions (woodwork for making musical instruments in India, shorthand and typing instruction schools). That is to say, each story is an independent experiment and is not connected to the other at all. But each offers a seed or point of departure.

    All this finally led to some more plotting on oral history concepts. Here’s the premise (naivety and ignorance on my part will also become evident!):
    — Each storyteller has many different stories.
    — Each story has many different storytellers.
    — Each story has many different theme and vice-versa, and so on…
    — There must be a way to connect a personal history with public events and places.
    — Each personal history can be connected to another using some categories/keywords/whatever-else: person, place (geography), timeline (when, year), event (is the story referring to an event like say, the SF earthquake of 1989 that someone else may also talk about in a seemingly unconnected way?).
    — A story may have weak or strong associations to a set of attributes. This association can be weighted.
    — There is an incredible range of stories recorded in university archives and other stand-alone oral history projects. The point is not to re-store them in a new space on the web. But there must be a way to re-point to them in a new scheme. That is, make the connections explicit.
    — There is an incredible range of stories among all of us that feel like they are so irrelevant and unimportant to a wider audience. Can we make them relevant?
    — Have multiple ways to add stories: upload, relink to existing, give contact for a volunteer to reach out, call in (toll free number), visit a recording center.
    — Map out origin of each story. A map with dots shows where a particular thread has received stories from, providing an incentive to fill geographical gaps.
    — Each story gets multiple keywords attached to it, reflecting the themes that the story touches on, besides the main subject matter thread.
    — Cross links: The keywords are a point of entry into the many storytellers-many stories paradigm.

    What’s the point?
    Let’s say something like this exists. Call it DH-Mosaic. Let’s say there’s a sociologist out there who is researching the history of wine making in the San Francisco Bay Area from 1850 to 2000. She has access to the usual set of archives at universities, and the state/national archives. She has access to archives of newspapers, books, papers, and such. And she does field work that gets her access to families in the industry. And she has access to DH-Mosaic, and she logs on and searches this clearinghouse/repository. Meanwhile, a few years ago, someone who was researching the history of high energy physics in the San Francisco bay area put out a book, and he had raw unused data. One of those pieces was an interview of a technician at a high energy physics lab in the region who mentioned that his family used to be in the wine making industry in the early 1980s and he left it in search of better opportunities. The high energy physics research didn’t have the space for this story. But the raw data is on (or linked on) DH-Mosaic, logged with metadata that points to: high energy physics, 1980, wine history, <name>, janitorial, magnets research, San Francisco bay area, Menlo Park. Our wine history researcher comes across this, and gets in touch with the author/owner of the story. In the course of talking to a new source, she finds out more first-hand sources about say, a labor struggle in the wine making industry in the 1980s, which takes her research in a slightly different direction….Similarly a reporter investigating the domino effect of the Vietnam War protests in 1970 in California may find a story and follow up with <name> after listening to his story of clearing up broken glass following the 1970 Stanford student protesters.

    What are some questions?
    — How to make these links?!
    — If you have a story, how do you define its relevance to a place/event/timeline/person/theme?
    — Who defines the relevance?
    — Within a specific story, how do you weight different aspects discussed?
    — Who defines this relevance?
    — How can this be done easily for a user interface?
    — Is it possible to get users and researchers to shape the categorization?
    — How do we get raw unused data (before it becomes a produced story/research paper) that just piles up for every researcher/reporter become a relevant trail of history?
    — There is the larger issue of copyright, censorship, spam and what-not. But I am not even going there yet :)!


    Loved this little quote from Michael Ondaatje’s book, Divisadero: “For we live with those retrievals from childhood that coalesce and echo throughout our lives, the way shattered pieces of glass in a kaleidoscope reappear in new forms and are song-like in their refrains and rhymes, making up a single monologue. We live permanently in the recurrence of our own stories, whatever story we tell. ”



  • “Mapping Segregation: Racial patterns of residence, land ownership, and rentals in Rivanna Dist., Albemarle County”


    My current research project is aimed at locating the residence of 6500+ inhabitants listed in the 1940 census of Albemarle County.  This work will be presented at the Virginia Forum in March 2014 and I am hoping that discussions of mapping techniques and relevant historical issues at ThatCamp can guide me in completing a project that will have technical and historical merit.

    There will be a dot on the map for each person that will display their 1940 census record when clicked.  I plan to group residents into four classes: white land owners, black land owners, white renters, and black renters.  I want to represent this information with semi-transparent spatial density layers created from the dots for each grouping, which ought to allow the viewer to intuitively perceive the spatial relationship between each group and a base layer of land ownership by race.

    The geodatabase will contain all significant census data for each individual.  This will allow map users to explore the relationships betweThatCampMapen categories of census data and land ownership by way of Boolean queries.  For example, it might be interesting to look at the relationship between race, level of education, and acres owned.  The query results can be exported to a spreadsheet for analysis and will be displayed as selections on the map (see map image and use Firefox to view interactive map at ).  At this stage the goal of this project is not to answer questions about racial segregation in a rural setting, but to provide data and tools to begin to formulate interesting and relevant questions that may emerge from viewing data in a spatial context.

    From a broader historical perspective, I have been interested in the formation of rural African-American communities after the Civil War from the enslaved communities that existed on farms and plantations.  How did freedmen acquire land and what was the quality of the land they purchased?  When, how and where were black churches and schools established? These are the institutions that formalized the identities of rural African-American communities.  When and how were the names of these communities incorporated into land and tax records? — in essence, recognized as distinct entities by white county administrators.

    My guiding concept in this research is that segregation implies not only social, but physical separation.  Therefore, it seems that spatial mappings of race and residence must be a foundational tool for studying themes like the one proposed for the upcoming Virginia Forum, “the creation, maintenance, or transgression of racial boundaries.”

Skip to toolbar