data mining

This paper spins from an analysis of several works of critical writing in the field of digital poetry, which have been documented at the ELMCIP Knowledge Base (http://elmcip.net).
The first stage of the research process has been trying to understand which tags have higher frequency, as one observes how the field has evolved in terms of theoretical analysis (content type: critical writing) and its correspondent classification in ELMCIP. Secondly, the database has been filtered in terms of critical writing tagged with the taxonomy “digital poetry” (higher frequency, 109 records). From here onwards, a period for analysis, 1995-2015, was selected, as it signals the twenty-year span of Web-based digital poems. Hence, the publication types selected were monographs and PhD dissertations tagged with “digital poetry” or having “poetry” in their title. This process created a smaller corpus for data analysis, mining and visualization. The selected works of critical writing include, amongst others, Hartman (1996), Barbosa (1996), Glazier (2002), Stefans (2003), Engberg (2007), Funkhouser (2007, 2012), Johnston (2011), Rosario (2011), Naji (2012), Dupej (2012), Emerson (2014) and Carpenter (2014).
The two questions I posed then were: 1) Is there a prevalence of self-referenced creative works in critical writing? 2) Is there a set of digital poems being more referenced than other?
By exporting the tables of cross-references from the ELMCIP KB into visualization software, I try to better understand and debate, by means of macro analytical visualizations and network analysis, these questions, following Franco Moretti’s notion of “distant reading” (2005) and application methodology in the field of electronic literature such as those developed by Jill Walker Rettberg (2012, 2013) and Scott Rettberg (2013).

(Source: ELD 2015)

The ELMCIP Electronic Literature Knowledge Base (http://elmcip.net/knowledgebase) is a human-edited, open-access, contributory Drupal database consisting of cross-referenced entries describing creative works of and critical writing about electronic literature as well as entries on authors, events, exhibitions, publishers, teaching resources and archives. The project has been developed by the Electronic Literature Research Group at the University of Bergen as an outcome of the ELMCIP project. All nodes are cross-referenced so users can see at a glance which works were presented at an event, and follow links to see which articles have been written about any given work or which other events they were presented at. Most records provide simple bibliographic metadata about a work or event, but increasingly we are also gathering source code of works, PDFs of papers and dissertations, videos of talks and performances, and other forms of archival documentation. While our first priority in designing the Knowledge Base was to provide a basic open-access online research infrastructure for an emergent field of scholarly and creative practice, providing researchers, teachers, and students with easy access to works, critical writing, and the context of a field, we are increasingly realizing its value as a base for further research in its own right. The Knowledge Base provides us with a growing pool of data that we are beginning to analyze using visualisations, social network analysis and other digital methods. This panel will consist of presentations of research developed by using information in the Knowledge Base as the basis for what Franco Moretti refers to as “Distant Reading” to better understand the discourse of the field and the works it encompasses. In this approach, instead of analysing individual works, we search for patterns across the entire field of electronic literature. The panel will present four different approaches to using the Knowledge Base to collect specific types of information related to objects, networks and practices of electronic literature and use digital methods to reveal patterns and trends from within the collected data that will hopefully inform a better understanding of specific aspects of the field.

Today, where information is continually transferred in the form of data, the word “information” has all but been exchanged for the word “data.” This shift of terms has aided in effectively transforming the world into a network-world of data. In many areas, and for many professionals, condensing information has become an almost exclusive preoccupation. This need to condense information through selecting and summarizing events—via the use of statistics, infography, visualization software, reports, databases, and animations—has dominated our mental landscape; it dominates the way we structure our perception of reality. Therefore, it is important to rethink what this phenomenon represents and how artists are responding to it. In this network-world of data, spam (which is unsolicited e-mail or electronic data sent en mass) has become one of the symbols representing the flux of disinformation, and/or unsolicited, information. Anti-spam is, therefore, a method of eliminating and screening the source data, a tool I call impedance. If we apply this point of view to contemporary art, we could consider the works of Pavel Braila, R. Luke DuBois and André Sier as anti-spam filters that allow the detection, screening, elimination, and subsequent reinvention of existing or non-existent data. In this essay, I propose considering the fundamental aspects of data mining, data visualization, projection mapping. From such considerations emerges the ability to generate, process, and recreate data from the work of these three artists. Finally, I introduce the perspective of the artist as a data miner, this reinvents its source in the new visual, social and political datascape. (Source: Author's Abstract)