User Tools

Site Tools


lauliac:solution_homework1_exercise1
Operational view

The graph presented here is always kept very simple. All starts with a single node, when a search is done (eg. “Duke Ellington”, but basically any music producer/artist).

One can then expand the node (left click ⇒ expand) and discover similar producers/artists. Created nodes can anew be expanded. There is no colors, only black shades, all nodes have the same size and shape ; thus the graph is very legible. However, when too many nodes are presents on screen, things are getting a little messy. Notably, the labels attached to each node overlap from time to time and hide the edges.

One can delete a node at any time (left click ⇒ delete), keeping only the most interesting nodes on screen. It's important when the user is doing a long session of artist discovery. The vizualisation feels clean and organic.

Articulatory view

The distance between a node and another indicate clearly the strengh of the link between two musics producers. The force-directed algorithm used here is very relevant, as the graph is repositioning itself as new nodes are created.

The interaction with the graph is simple and efficient. Two clics are necessary to expand a node, maybe one would be preferable but it's still very usable ; the user is never lost.

It seems impossible to move the whole graph on screen, thus making impossible to access the farthest nodes ; that is not pratical;

Semantic view

This vizualisation is aimed at simplicity.

The way this graph software invite the user to discover by himself a whole universe of music artists is very attracting. However, there is very few data presented here. The graph fulfill the bare mission, discovery of new artists, but another support (such as Amazon, Last.fm, etc.) is needed to “discover” really the artists. As we can see in the bottom-right, the vizualisation is sponsored by Last.fm, the well-known music tracking web service, and this graph match well the business.

Some information that could be presented along includes: music styles and genres (eg. by using node colors), artist biography and historical data, specific nodes for music pieces.

/net/html/perso/melancon/Visual_Analytics_Course/data/pages/lauliac/solution_homework1_exercise1.txt · Last modified: 2012/02/29 23:17 by lauliac