thoughtshuffler is a new way to read, search, or blog, on desktop or handheld.
iPhone mock-up
visualizes a user conversing with a text; also in iPhone viewable version.

THOUGHTSHUFFLER software is a unique user interface and a proprietary heuristic. It offers the user unusual efficiencies for:

  •  submitting queries and reviewing results from search engines
    (text, images, video)
  • reading and filtering blog posts and comments
  • drilling into large documents for parts of greatest interest and avoiding the rest
  • quickly “groking” the relevance and value of the content to a user’s needs.

The software delivers these benefits to displays of all sizes, from desktops to hand-helds.

The software interface of THOUGHTSHUFFLER is a simple, spreadsheet-like format that provides the user with simple clicking (or tapping) gestures to perform complex content manipulation functions. The interface allows a user to:

  • submit search queries, quickly review results large sets of results, and make rapid query refinements
  • use pre-existing meta-tags to quickly filter and find images or video
  • sift through postings and comments on blogs to display only those of interest, creating focused “dynamic threads” from hundreds of thousands of comments
  • drill down into a document to quickly get the overall gist, or to focus on a sub-topic defined by the user
  • juxtapose similar texts and enable a fine-grained, easily controlled “compare and contrast” content display
  • see the context of every instance of a phrase or combination phrases
  • display content that represents the summary of a page or sub-topic or full document
  • move through hierarchies of detail in the document.

The proprietary heuristic behind THOUGHTSHUFFLER is used to automatically extract a carefully determined cluster of tags from full-text content, which renders the interface functions mentioned above with fewer user interactions and greater speed-to-relevance for the user. The heuristic uses a theory of how cognitive systems form distinctions and create structural “coherences” in the course of understanding new input. As a result, the user is offered:

  • multi-dimensional tagging, which overcomes some of the limitations of flat taxonomy schemes
  • suggestions for important phrases that represent the range of topics covered in a document
  • key content fragments that come closest to summarizing an entire document or any subset based on key phrases.

The THOUGHTSHUFFLER prototype runs entirely as a service and is available for live demonstration via remote server to IE7 or Safari browsers.


self-describing movie of first UI, uses thoughtshuffler to describe itself; just follow the bouncing cursor.

first experiments with images is an illustrative, but sloppy, example of a thoughtshuffler UI for images.

movie of desktop browser version runs a canned QuickTime movie of LIVE CODE from an earlier, clunkier UI.

walk-through of next-gen concept [PDF] uses annotated sketches to give a walk-through of the concept

HTML thoughtshuffler described in a text with non-annotated screenshots.

thoughtshuffler history shows a sequence of prior experiments, including authoring.

live thoughshuffler demo [in the process of being revived]

    email to pan @ this domain.