Use the RSS data with tags from del.icio.us or from the previous homework and write a SPARQL query for it to produce a "Tag cloud" in KLUCIS - a rectangle filled with tag names in various fonts.
Font size is proportional to a rounded logarithm of the frequency of that tag. E.g. tags, which are used 1 time could use font size 8pt, 2-3 times - 10 pt, 4-7 times - 12 pt, 8-15 times - 14 pt, 16-31 times - 16 pt, etc. The calculation of the font size could be either added to the Velocity Context by some custom KLUCIS component (e.g. you can introduce a class, which inherits from a rectangle called "TagCloud"), or is calculated inside a Velocity template by some advanced "math" object.
One challenge is that SPARQL does not provide aggregate methods, i.e. way to count how many times some tag has been produced. This should be counted by some iterator in your Java code, which you can add to the KLUCIS standard distribution.