Part I. Toward a smarter data science. Why textual data from the past is dangerous -- From fantasy to engagement -- Words are keys and words are barriers -- Critical search: a theory -- To predict or to describe? -- Part II. The hidden dimensions of temporal experience. The many windows of the house of the past -- Of memory -- The distinctiveness of certain eras -- The measure of influence -- The fit of algorithms to temporal experience -- Whither modernity -- Attacks on environmentalists in Congress -- Part III: Disciplinary implications. A world map of world culture, purged of bias -- The future of the art.
Summary:
"The Dangerous Art of Text Mining celebrates the bold new research now possible because of text mining: the art of counting words over time. However, this book also presents a warning: without help from the humanities, data science can distort the past and lead to perilous errors. The book opens with a rogue's gallery of errors, then tours the ground-breaking analyses that have resulted from collaborations between humanists and data scientists. Jo Guldi explores how text mining can give a glimpse of the changing history of the past - for example, how quickly Americans forgot the history of slavery. Textual data can even prove who was responsible in Congress for silencing environmentalism over recent decades. The book ends with an impassioned vision of what text mining in defence of democracy would look like, and why humanists need to be involved"-- Provided by publisher.
This resource is supported by the Institute of Museum and Library Services under the provisions of the Library Services and Technology Act as administered by State Library of Iowa.