distant reading
Graphs, Maps, Trees
On 21, Dec 2015 | In Picturebooks | By Chris Vitale
Moretti, Franco. Graphs, Maps, Trees: Abstract Models for Literary History. Paperback ed. London: Verso, 2007. Print.
Referrer: Matt Gold
Categories: digital humanities, distant reading, data visualization, historiography, methodology
Annotation:
This essential piece of the Digital Humanities canon outlines the efforts of Franco Moretti and his team in their journey to distant read the literary work produced over a few centuries. Graphs, maps and trees are shown to be valuable tools for literary scholars, like their natural and social sciences counterparts. The concept of distant reading, application of an analytics and quantification driven approach to reading wide ranges of texts was groundbreaking for the literary community. This historiography of the literary canon illustrates the merit and value of distant reading as a legitimate methodology. Acting as a roadmap for how quantification and visualization can augment, complement, and completely topple traditional research methodologies, this text is a fundamental piece of this Picturebooks project.
Computational historiography
On 21, Dec 2015 | In Picturebooks | By Chris Vitale
Mimno, David. “Computational historiography: Data mining in a century of classics journals” Journal on Computing and Cultural Heritage (JOCCH) 5(1), 2012
Referrer: Scott Dexter
Categories: digital humanities, distant reading, data mining, computer science, methodology
Annotation:
David Mimno follows in Franco Moretti’s footsteps in an effort to data mine a massive archive of classics jorunals. This distant reading is the preferred methodology for Mimno who has identified the ability close read such a wide array of documents as unrealistic. The paper discusses the use of computational tools that allow for the statistical analysis of the corpus. The work is explicitly complimentary to traditional scholarship. The collection that Mimno is working with has been OCR-ed from over twenty classical philology and archaeology journals. Outlining the tools used in statistically driven mining of texts, Mimno discusses tokenization, removal of stopwords, word distance and divergence, and topic modeling. The algorithmic representations of these computational methods are given as well as an introductory discussion of the ways they work and are used. Finally, Mimno presents his findings in the forms of graphics, topic models, and observations.
Significant Themes in 19th-Century Literature
On 21, Dec 2015 | In Picturebooks | By Chris Vitale
Jockers, Matthew and David Mimno, “Significant Themes in 19th-Century Literature” Poetics 41(6):750–769, 2013
Referrer: Scott Dexter
Categories: digital humanities, distant reading, data mining, computer science, methodology
Annotation:
Matthew Jockers and David Mimno discuss their distant reading project regarding the themes of 19th century literature. To do so, the two researchers mine and model topics from over 3,300 works of literature. The literature encompasses British, American, and Irish texts. Taking a variety of other factors into consideration the two seek to find trends in the themes of novels from that century. The methodology used incorporates counting words, tokenization as well as topic modeling. The tool Mallet is used by the researchers as the main engine for mining the text. Jockers and Mimno discuss the preprocessing elements (stopword removal, Bag of Words segmenting of the text, part of speech tagging of nouns, and modeling of the topics). Finally, the analysis and observations are given. Topic modeling is shown to be a scalable solution for those interested in reading massive selections of hundreds and thousands of books.
The Digital Humanities Unveiled
On 21, Dec 2015 | In Picturebooks | By Chris Vitale
Spratt, Emily L. “The Digital Humanities Unveiled: Perceptions Held by Art Historians and Computer Scientists about Computer Vision Technology” (Self Published).
Referrer: Scott Dexter
Categories: digital humanities, distant reading, data mining, computer vision, art, art history, computer science, methodology
Annotation:
This paper outlines a survey completed by both art historians and computer scientists in relation to a computers ability to interpret aesthetic and beauty. The value of this work lies in the responses of this survey. Computer vision is rapidly becoming a more accepted and accessible method of examining art. For art historians and computer scientists, the implications are obvious. This digital humanities project used, “twenty-one questions for art historians and sixteen for computer scientists that were intended to shed light on field members’ knowledge of the capabilities and applications of computer vision technology, attitudes and perceptions about the use of it, and reactions to the meaning of this type of digitization in the humanities.” Spratt discusses the positive and negative reactions to computer vision’s ability to detect and automatically recognize aesthetic experiences of beauty. Channeling philosophy, Spratt defines what these variables mean for her survey.
Toward Automated Discovery of Artistic Influence
On 21, Dec 2015 | In Picturebooks | By Chris Vitale
Saleh, Babak, Kanako Abe, Ravneet Singh Arora, and Ahmed Elgammal. “Toward Automated Discovery of Artistic Influence.” Multimed Tools Appl Multimedia Tools and Applications (2014). Web.
Referrer: Scott Dexter
Categories: digital humanities, distant reading, data mining, computer vision, art, art history, computer science
Annotation:
A team from the Department of Computer Science at Rutgers experimented with art and computer vision in 2014. Using advanced computer driven recognition of images, the team was able to explore the influence of other artists on particular pieces of art. The study was focused on two types of computational inquiry: “discriminative vs. generative models” as well as feature extraction and comparison. The importance of this work is the argument that computers have a, to a certain degree, level of ability in recognizing the influence of other artists in multiple works of art. The work of a Art Historian is arguably completed by an automated process that involves computer vision and classification. The high level computer algorithms are paired with art history style analysis of the paintings similarity to compare the traditional and novel methodologies. This paper is a valuable source of tools and methods for computer vision in illustrations.