Category Archives: Databases

CARTODB  The Power & Limits of Perspective!

In Women’s Studies, we examine concepts from different points of view.  Engaging in the idea that ‘it matters where you are standing’, because what you are experiencing shapes your interpretation of the problem or context.  Indulging in the examination of a variety of perspectives does not always lead to a conclusion or an answer, but it imparts a fuller and more complex understanding of the issues being addressed.

CARTODB opens the opportunity for novice geeks to play with data and visualize information from a different perspective.  The key word here is “visualize”.  Mapping allows everyone!  (if I were writing a grant I would say, ‘the general public as well as scholars and students’); to “see” human experience in new configurations.  Ideas are sometimes hard to follow, but evidence for an idea can be provided by demonstrating connections and patterns and “placing” the idea onto a map.   A created map is “a map” which displays information, but CARTO has widgets for interactive time sequences, pull ups for additional material or explanations for aspects of the map.  The process of building the map, plugging in ‘layers’ (multiple data sets) of data into a ‘corpus’ (total collection), is exciting for the researcher because it allows them to “play” with ideas!   For example, in analyzing the Slave Narratives using different map views (point identification, “heat” (concentration) or time sequencing) reveals different aspects of the Slave Narrative project – its limitations and biases as well as providing a great deal of information on the slaves and the interviewers.  This type of mass collation and presentation preparation would take years without the speed of computing.  But the computer cannot interpret the meaning of the data.

Mapping allows for the interpretation of large quantities of data, and the consultation of a wider scope of resources. (Tom Hitchcock “Historyonics”).  The caveat being that the data has to be in a digital format, this is a limiting factor and the limitations of whatever ‘data set’ is being used should be recognized.  Sharon Musher (in Oxford Handbook of the African American Slave Narratives) makes an excellent and pertinent reminder of how even a seemingly large and comprehensive set of information can be misleading in its content.  Great questions that mapping answers are locations, patterns and relationships (who is where and when and how does that change over time).  However mapping allows us to see the significance of these relationships in new ways.  As in Digital Harlem, where the placement of a patrol officer on a street corner and involvement in traffic accidents can now be understood as intersections of racial politics. (Robertson).

A bonus is that skeptics of the value of history (like Tom Hitchcock) will have something concrete, built on numerical manipulation that they can “see” as evidence, because ideas may seem too fuzzy for them to comprehend.  CARTODB is a fabulous tool for allowing the explorer of an event, a neighborhood, the spread of ideas globally, the impact of words, or as in the Slave Narratives…the impact of human actions, individually and communally, from more perspectives.

Make no mistake, working with CARTODB is interfacing with computing.  It requires patience for the tedious and repetitive steps of data entry, learning the “tools” and patterns of operations within the program and is never as ‘intuitive’ as the newbie wants it to be.  After my initial foray, I can see, as with my phone and Voyant, the more the platform is used, the easier, more fluid and of course faster the user will become.  I just keep praying that eventually the very act of learning this type of tool will become more fluid.  If you are a person that needs to understand the overall picture of the tool, some time spent in on-line helps and guides is worthwhile.  If you are a person who is a digital native or who doesn’t get frustrated by getting stuck and starting over, dive right in.

“Seeing” the Slave Narratives with Voyant

Voyant Tools  is a web-based text reading and analysis environment designed to facilitate reading and interpretive practices for digital humanities and the public.  http://voyant-tools.org/docs/#!/guide/about

Voyant Tools is an entry point – software that allows you to do basic computer-assisted analysis of data.

Text or edited and prepared data can by studied and analyzed in new ways and ‘plotted’ so that it can be saved and shared with others.   Using Voyant allows others to explore your data through interactive panels that can be inserted into online publications and presentations.   Voyant can also be used as a development platform to design new forms of analysis and visualization.  The main visualizations are: word frequencies, word clouds and “distinctive words”.

I am also including a link to the scatterplot for this corpus.  Clicking on the “Analysis” option > Principal Component Analysis provides  further evidence of the significance of “race” as a feature of slave experience.   http://voyant-tools.org/?corpus=2d0c91a1da599627fea6d35d607795a0&view=scatterplot&stopList=keywords-96814a63d28a5e281f75e02c891dcdbe


The trends graph indicates the significance of location.

I found Voyant to be an interesting tool which will be of great use when texts are clearly comparative.  It is great for collating data!  It provides the ability to analyze text to see trends in language and usage.  The word cloud is interesting and illustrative.  In the Slave Narrative data set on which I was working the common words were Old (which I ignored as the subjects of the text were elderly), Come and Time.  Whereas I would have expected more direct slave references which were secondarily popular.  The nature of the documents (by state) also provided insight into regional variations, but also made me recognize the states from which the largest number of narratives was drawn, would bias the representations and conclusions.  Understanding the Data (Musher, The Other Slave Narratives) is critical to recognizing the significance of patterns.  Having a larger context besides the quantitative data is essential for interpretation.  Also it became evident as I was working with the Data, that my choices were subjective and these choices would also “skew” the results…my lack of technical prowess also would play a role in the accuracy or distortion of the results.

For the new user….agh!  I am sure this tool would be easy for humanists with digital strengths.  For the digitally challenged, it is probably pablum, and a good entry point to Digital Mining and Modeling tools…but we are entering a “Brave New World” where the faint hearted must learn to persevere.  I believe spending some time in the on-line guides would be very helpful.  I played around the examples and information on Voyant’s website, but before attempting a larger project – I would try to learn to manage and navigate the tool better.  Time being an issue, I could not do that.  My simple tasks became time consuming, because if I tried to backtrack or change some aspect of the data my progress was lost.  Trying to return to old screens or navigate between the tools was not as “friendly” as I hoped.  Transforming from a historian who uses traditional sources of evidence to build up thesis and wider connections – I miss the library shelf (Underwood, Theorizing Research Practices).   it is a dramatic change to reverse and drill down into the minutia of checked boxes, single characters and itemizing the search.   The good new is that when you get interesting graphs or can make unexpected connections and conclusions…ya feel great!

“Learning is not attained by chance it, it must be sought for with ardor and diligence.” Abigail Adams

A review of the American History & Life Metadata Search for Popcorn

A search for related information about my Whirly Pop popcorn maker yielded:  Popped Culture: A Social History of Popcorn in America” (Book)

Bowl of popped corn
Wikimedia Commons

It is very important that the Metadata mentions after the title that this object is a book.  Since I searched for popcorn…it might have been anything.

The American History & Life Fields included the Author (Trudy Eden), the Source  (H-Net Reviews in the Humanities & Social Sciences. Jul2004, p1-6. 6p.); The Historical Period (1830 to 1999); The Document Type (Book Review) ; Subject Terms POPCORN, NONFICTION, SOCIAL change; Keywords (Andrew F. Smith important as he is the author of Popped Culture),  Reviews & Products (This took me to another review by Philip Hiscock in Ethnologies), Abstract, Notes (incl. Publication Information: Washington: Smithsonian Inst. Pr., 2001. 264 pp.), Author Affiliations (Department of History, University of Northern Iowa), ISSN (1538-0661), Accession Number (15099661)

This Metadata did not include a Description or Summary Field that would have given me an idea about the review.  Although the Period was identified, the context of the object was not.  A “Review” for social change is not the same thing as “Review” for ethnography (which only came from knowing the nature of the publications for the different reviews). The publication information did not let me know if the book was drawn from an exhibit or independent research on the topic – which might have been included as part of the exhibit).

I appreciated the embedded links made searching the related sites and information quick and easy.

The Author information and Reviews and Products information made it possible for me to “place” this review in some of the fields of history to which it belongs (ethnography and Social change).  It did not tell me if the reviewer is a significant person in the field.  The subject field was very limited as it did not lead me to any other potential resources or search terms for my Popcorn maker – like Cooking or any information about the location of the subject other than (America) or its connection to industries like theaters or circuses.  It did not identify related materials (if it was in an exhibit) or if this book covered the entire history of popcorn or just a period of popcorn or if this was the epitome of Popcorn and there has been a decline since.

My search of the metadata of the American History & Life reinforces my conclusions described in my Search Review.