Feb

3

The Weathering of Words or All-Weather Tiers

Filed Under Told You So 

Andrew Kuo NoAge Santos

“Google Graphs.” Does it exist? No. Should it? Yes. Why? I hope you’re sitting for this one. (Really. Sit down; get settled. This post is a long one.)

Anyone who knows me, or has been in my company for somewhere between a dinner party and an e-mail-to-phone-calling-scheduling-for-coffee length of time will no doubt have to endure one of my diatribes about the organization of information.

Organizing information is not only the hear and now, but it is the future. (Please refer to “Data Analysis,” section 3.5, page 180, of the reprotext found below your seat.)

If the twentieth century was about the proliferation and access of information, than the twenty-first century, must be about the organization and analysis of that information. Simply, if the inventive quality of information is saturated in a modern era, than the analysis of the information is not; especially when there is so much to research and cross-research in a multitude of ways.

I first learned about DocuBurst from an article in The Toronto Star about two years ago. An interactive and graphic index, DocuBurst is the brainchild of Christopher Collins who, along with, Sheelagh Carpendale, and Gerald Penn, developed the project as part of his PhD research at the University of Toronto. In Collins’s terms, DocuBurst (document + sunburst), is a tool to help “visualize document content using language structure.” (Refer to “Semiotics and its Discontents” in the Notes section; located at the back of the reprotext.)

DocuBurst’s function is two-fold. One, it provides the frequency in which a searched item appears (like a regular index). Two, it visually displays an overview of the magnitude of the searchable item within context of the document content. It does this by highlighting desired item searches, including words that occur before and after the searched item, and by a radiating pinwheel graphic, which bridges “is a” relationships to the searched item, rather than, “see also” relationships, (think related, not relative) to determine detailed and relevant information with each subsequent layer of burst. The interactive aspect of DocuBurst allows the search to occur at any point in the document (the index of the index of the index) while colour coding quickly shows words connected to the searched item.

If DocuBurst’s explanation sounds involved, you should see it’s pinwheel information graphics. Specific searches tend to lead into stunted sun rays, “bursts” away from the central search, thus commenting more on the DocuBurst itself, rather than searched item. (See, Spiro Novak’s essay “Rays Away: How Smart Devices Steer from Searches to Solstices,” Farrar, Straus and Giroux, 1999, in the Bibliography of the reprotext.)

In order for the graphic organization of information to be effective, it must provide a general overview, not a detailed one. Specifics potentially can take away from the overall picture. The purpose of the project, defeated, entirely.

Wordle, “a toy generating ‘word clouds’ from provided text” provides a simplified, general picture, through the use of big, bolded type. Although more like a Dada poem, both in the words the “toy” chooses to extract from furnished text, as well as how it chooses to position words, with and against each other, Wordle’s emphasis and de-emphasis is an opportunity of representation than presentation as it allows the viewer of its clouds draw causal relations, rather than conclusions, between the missing gaps. In essence, it includes the element of interpretation, much like Google where relevant searches are determined by popularity. That is, a majority of shared interpretation.

Imagine now, if Google shared DocuBursts’s specifics, with Wordle’s suggestion, each time an item was searched. What you would get is an enhanced form of relevancy as users could zero in on sections, not sites, for information, through a colour-coded, simple, graphic which indicated the usefulness of the content before investigating it in greater detail. The graphic would be in the place of the traditional listing Google provides, or even in addition to users who have the GooglePreview extension–where in addition to the listing of the search, a thumbnail of the website featured, is also displayed. More searches then could be displayed on a web page, quickly displaying an even larger, general picture, of what sources stand out among others.

Perhaps, instead of Google Graphs, a more accurate name for the function would be “Google Forecasts.” An immediately recognizable visual index pointing to the likelihood of desired information, with a small gap of unpredictability and probability of interpretation. (See Andrew Kuo’s charting of music albums and concerts reviews at Earl Boykins blog, under References, in reprotext.) It could very well be, the next, “now” thing.

NB: All references–names, titles, essays and sections–included in the above blog post are fictitious with two exceptions. Farrar, Straus, and Giroux is an imprint of Macmillan. Also, Andrew Kuo’s blog is an amazing place to get your b-ball, hip-hop, and Excel graph on. The referred reprotext mentioned in the blog post, does not exist. The author apologizes for any inference to references of same or similar information, unknown at the time of writing and publishing this post. The author welcomes corrections, amendments and indexes to similar, supporting information.

IMAGE | Andrew Kuo | No Age at Santo’s Geographically | 11 November 2008 | Excel

Comments

Leave a Reply