This week I was busy preparing another reflection tool for experimental use. This time it is called ReScope and a little tool for self-reflection based on the personal tagging habits on del.icio.us. The second prototype is now available for live-use at my experimental server at the OUNL. Those who check my homepage, will recognize the look of the tool from there. In fact, the version on my personal homepage was the first prototype. The current version is a complete rewrite of the first approach and will address the needs of scientific testing.

The pre-version of ReScope was mainly as a show case about how to encode multiple information dimensions into a single tag cloud. During the last year I watched carefully the development of the tag cloud and found some interesting use cases how such a simple tag cloud may stimulate reflection. I wrote a short paper for a special track at ICL 2008 in Villach (Austria), for which I invented the name. In this paper I report on the first experiences of using such a tag cloud in practice. Although, there is no hard evidence about the actual benefit for reflection support, I discovered some interesting practical implications from the early design study. However, it is time to extend the first impressions and ground them on serious empirics. The only problem with respect to experimental studies is the lack of flexibility of the first study: as almost everything was hard coded for a single user setting, scaling was a critical demand for the next steps.

The Core Idea

The core idea of ReScope is to expand the information that is encoded in a tag cloud, so it becomes meaningful to a user within a learning process. This should help people to become more aware about their learning processes. ReScope supports this by relating the overall usage of tags with their recent use. That way the tag cloud becomes a dynamic structure, instead of a growing pile of words.

This idea is quite different from the mainstream in tag cloud research, where the main focus with regard to tag clouds lies on collaborative semantics, search, and indexing. One important problem that arises from this focus, is that many users fail to identify the benefit of a tag cloud for themselves. This problem has been tackled by Moritz Stefaner in his Master's thesis. Where he focused on aesthetics and design factors of tag clouds, the work on ReScope seeks to identify the factors and relations between tagging and learning.

New Features

For the new version I abandoned all code that was based on self-made modules for higher level functions and used the latest mootools javascript library.

This version now implements a JSON script loader through which it is possible to load the tag cloud information directly from del.icio.us. This is has the great advantage that the second prototype ReScope still works without a dedicated server component, but allows more dynamic content generation.

Then I added a little piece of code that allows ReScope to use the settings of users of my team.sPace system. That way a user who is already registered in team.sPace can directly get a personalised tag cloud once they are logged in.

For those users who are not logged in, it is possible to type their del.icio.us nick name in a small form. ReScope then fetches all the information for that nick.

Finally, for those people who like to bookmark their personal tag cloud without registering, I added a hash extension of the URL. The client side code is aware of the hash tag and uses that tag as a del.icio.us nick name. UPDATE: For easing bookmarking I added a link to the UI that allows users to bookmark their tag cloud.

ReScope now works on all major web-browsers.

More to come

For the experiment some features are still missing. One feature is the micro-blogging tool that allows registered users to make notes. This tool will store the visible tag clouds alongside with the note, which will then enable the users to go back in time, and see not only the notes, but also the state of their web-readings at that point.

As soon this component is finished, we can start inviting the people to the experiment.