Trimming vs Flat

This nature article reports concisely on  an alternate design for the back button history stack. The advantage of a flat recency ordered list comes at the cost of losing a rudimentary hub mechanism.

The trimming process favors home pages or google result sets.  Mozwho aims to explicity account for these with the notion of hub detectors.
|||
Posted at NaN:NaN

Observer Progress

The latest updates in CSS maintains a core array of VisitObjects.  For each tab, there's an array of pointers into this visit object.



Both traces are available in dropdown menus.  In addition, both dropdowns feature a base tree transplanting operation. After visits to "A--B--C" followed by a back operation in the custom menu to return to  A, a popout is created from A containing "B--C".

Additionally, a UI placeholder for the associative tagging/drop ops has been included.
|||
Posted at NaN:NaN

isVisited()

Hmm, the history datasource (?) has a isVisited method -- definitely better than the search operation I'm currently doing for base decisions.
|||
Posted at NaN:NaN

Spaces

Getting the basic logging structure in place took longer than expected, but it works more reliably now, catching one page tab views.

Next, I'm thinking about project spaces.  As a quick hack to explore the usefulness of this datasource in absence of a strong persistence layer, I've implemented a temporary route using WDDX to serialize a windowHistory structure. 

The "portal" button launches a dump of your Window history. The first section is global, the second each of the tabs. Hovering on a link in the first area will do a properties dump in the embedded iframe.
|||
Posted at NaN:NaN, Published in: Code

Foraging

A nice expose on academic information foraging -- a good use case for mozwho.
|||
Posted at NaN:NaN

Page Transitions

The method (ex. UI trigger, web page link text) of page transition seems indicative of the relationship between to pages in the history sequence.  How to operationalize this?
|||
Posted at NaN:NaN

Infinite Serialization

A fresh 0.0.2 xpi is up which has a working hierarchical back button, but only for jumps initiated from the menu.  The first xpi suffered from a quick fix deletion of a pointer into the menu in the history object.  The menu pointer apparently caused my serialization effort to try and xml the whole dom, or at least some recursive component.


|||
Posted at NaN:NaN

25% of users customize their homepage

A recent First Monday pub suggests that 1/4 of individuals change their homepage, typically to a site in the top 3 most visited. Almost 40% of the remaining have a portal as their homepage but tend to not use it. 
|||
Posted at NaN:NaN

Tangle

Tangle captures some of the same information we're playing with in Sextant, but in a multi-user collaborative fashion via a proxy server implementation.  Check out the PDF article for details -- the essence is cataloging inbound and outbound linkage.
|||
Posted at NaN:NaN

bugzilla.mozilla on hierarchical history

Reported 1999-12-11, bug 21521.  Related re-file at 187410 and a look at the forward button at 187187 |||
Posted at NaN:NaN

Browsing Icons

"Browsing Icons" provides some prior-art to reinstating history context based upon current activity.  My visualization isn't quite as spiffy yet, but we're getting there:


Bookmap is a previous take on a inferring a best-of your surf trail with some snazzy fisheye history page thumbnail super-foo.
|||
Posted at NaN:NaN

ARC-ing

Yes, they are arc's in RDF and that's causing great pain right now.

A more fun ARC is "automatic resource compilation".  This www7 work  evaluates a particular method for inferring, what we call here, hubs.  While we can get a lot out of page following link analysis, spidering the links on pages visited often/for long active periods is the next step.
|||
Posted at NaN:NaN

Full Text Work

While we're avoiding full text analysis initially, it is a fruitful area given the system architecture.  As recounted by J.Pitkow in his 1997 GIT thesis:

Letizia manages a fequency based keyword profile for each user based upon visited Web pages. When a user is visiting pages, Letizia searches the pages connected to the current page and uses the keyword profile to determine relevancy.  Recommendations are then made to the user about which pages may be of interest.  This is more of an artful integration of ideas than a methodological contribution. p. 93.

In order to mimize impact on browser response time, MozWho will use a sequence of timeouts from page load.  If the user is still on the page and has shown some interest in it, or it matches in some way, further analysis of the page's links and content may be triggered.

It is also worth note, as also reported in this thesis, the top user reported problem in the mid/late 90s was finding known pages.

|||
Posted at NaN:NaN

Link Appearance

Another fruitful are for online activity is augmenting links.  The simplest approach conveys continuous variables recency of visitation, community ranking, or probability of interest. 

The probability angle was tackled by Chris Olston and
Ed H. Chi in  ScentTrails: Integrating Browsing and Searching on the Web.

Other types of link augmentation abound.
|||
Posted at NaN:NaN

Different Interests

A key challenge is separating a user's interest into distinct domains. JP McGowan, N. Kushmerick, and  B. Smyth offer some insight in Who do you want to be today? Web Personae for personalised information access.

|||
Posted at NaN:NaN

Publishing Trails

Mozwho is a rather opportunistically directed project.  Low hanging fruit are picked as we meander through data sets and UI constructions. 

One killer app that's been rising in my conversations in this space is publishing web pages as a result of browsing trails, an idea that I think can be appropriately credited to VB's memex. 

At Hypertext '03: Navigation, Path-centric browsing, Navigation metaphors, Directed paths, Walden's Paths, Path Engine.

Walden's Paths is a project to allow teachers to create guided trails for classroom instruction.  They use a framed approach -- perhaps a Mozilla add-on client would be a nice addition.

Try a path on functional programming, or Cervantes.

The ability to add commentary digested with the page in question is a meaningful distinction from even a well annotated link list -- our target functionality in MozWho.

|||
Posted at NaN:NaN

Compound Elements

There's also some nice work on identifying documents spread across URLs at HT03.  The term is "compound elements" and aims to facilate operations on documents, not pages.

K.McCurley's done lots of other interesting work, including the notable "Surfing the web backwards".
|||
Posted at NaN:NaN, Published in: Ideas

Zoomable Browse Maps

I've re-implemented the browse map concept in flash with mouseover feedback and onclick drill down to associated information: links followed, time on page, etc. This map shows the history of the whole window, providing a tab independent history.

The current long range goal is to make this available in the sidebar and use zooming to deal with screen real estate issues.  The discontinuities seen below, in between progressive nodes with no connecting lines, result from browser UI url  acquisition, not link following.

The browser, as opposed to page, acquired urls are likely to correspond to different tasks or at least different subgoals.  I'm excited about capturing this as well as about the utility of full window history functionality.
|||
Posted at NaN:NaN

Prior Spiral Art

The spiral approach to visualization is not new and it's applicability to temporal data is well argued.  Rhizome's used it to great success to show a temporal trace of the site's acitivity.  See also a PeterMe discussion and a 2001 InfoVis pub,
|||
Posted at NaN:NaN