Group items tagged

The HathiTrust Digital Library is a partnership of research institutions and libraries working to securely preserve historical collections to be accesible long into the future. These collections are open access, and include a wide spectrum of cultures across a variety of different time periods. The partnership has been recently engaged in legal disputes regarding alleged copyright infringement in their Orphan Works Project. In addition to basic access to many of the collections, the HDL offers search functions within the documents that allow for new uses of the texts, such as text mining.

Blogger whose identity I could only trace to as John has written into this Digital Humanities website. He shares with us an announcement that Google has now opened their text mining project that allows for better searching using frequency of words and phrases. This tool is compared to a game using a Star Trek example.

Marche's article criticizes digital humanists for a perceived failure to adequately address the human and interpretive nature of literature by treating it as data. Two core issues identified by Marche is that literature, unlike statistics, is terminally incomplete - that parts frequently are missing or shifting - and that data mining efforts fail to account for context in literature. Marche argues that current data mining efforts are flawed because "algorithms are inherently fascistic" and that "meaning is mushy." Marche does not oppose digitization efforts and in fact welcomes the translation of texts into digital formats, rather Marche argues that literary meaning cannot be as readily quantified as numbers - that "insight remains handmade."

This article reports on the ruling by Harold Baer, Jr. which held that the HathiTrust's mass digitization is fair use. The judge explained in his opinion that the HDL's project is not only fair use in and of itself, but that its potential for text mining and the facilitation of access for print-disabled persons are transformative in nature, and can serve an entirely different purpose than the original works.

In his blog, Dan Cohen decided to revisit a topic that was cover in the Chronicle of Higher Education. This data-mining related article discusses the issues with educational testing and growing technology in the humanities field. Devices that can browse an entire database of knowledge pin pointing specific facts. This device is then compared to the relationship between the calculator and math to this device and history. Just as the calculator has made memorizing certain mathematical principles pointless in testing, this device is said to make multiple choice test irrelevant for history. Similarly, cell phones, pdas, and tablets have been able to fill this gap already.

This is a follow-up article to a post Cohen wrote on Wikipedia and its relation to Google and Yahoo. In this post, he discusses the validity of Wikipedia as a tool to create text profiles of subjects for search engines.